66 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016
Vector Median Filters: A Survey
Roji Chanu1, and Kh. Manglem Singh2 1Department of Electronics & Communication Engineering, National Institute of Technology Nagaland 2Department of Computer Science & Engineering Engineering, National Institute of Technology Manipur
Abstract nature, nonlinear characteristics and noise corruption In this paper, a comprehensive survey on vector median filters to statistics. remove the adverse effect due to impulse noise from color Impulse noise is a high energy spikes having large images is presented. Color images are nonstationary vectored amplitude with probability greater than that predicted by value signals. Hence, nonlinear filters such as vector median Gaussian density model occurring for a short duration. It filters are more effective than linear filters. A number of is required to remove noise in the preprocessing stage to nonlinear filters are proposed in the literature. They have been categorized into 12 groups and discussed in details. prevent degradation of image quality.Many nonlinear Keywords filters have been proposed in the literature for removing Impulse noise, Vector median filter, Quaternion, Nonlinear, impulse noise. In this study, a large number of nonlinear Sigma vector filter, Entropy vector filter. filters are categorized into 12 families. This paper aims at summarizing the recent developments in the vector median filters for removal impulse noise 1. Introduction from color images. Section 2 describes the categories of vector median filters. In Section 3, a commonly used Color images are widely used on daily basis in printing, impulse noise model is described. Popular filtering photographs, computer displays, television and movies performance criteria for evaluation filters are given in and thus color image processing plays a crucial role in the Section 4 followed by conclusion in Section 5. field of advertising and dissemination of information [1]. The use of color images is increasing in medical images and remote sensing. Color information are being used in 2. Category of Filters many color image processing applications like object recognition, image matching, content-based image In this section, the commonly used filters used for retrieval, computer vision, color image compression, etc. removing impulse noise are grouped into 12 categories. [2]. Color images are corrupted by noise due to They are malfunctioning of sensors, electronic interference, 1. Basic Vector Filters imperfect optics, or fault in the data transmission process. 2. Weighted Vector Filters Noise introduces color fluctuation making pixel values 3. Adaptive Vector Filters different from the ideal values and thus produces errors which complicate the subsequent stages of the image 4. Peer Group Vector Filters processing process [3]. A filter transforms a signal into a 5. Fuzzy Vector Filters more suitable form for a specific purpose [4]. Filtering 6. Hybrid Vector Filters gives an estimate of signal degraded by noise. Since color 7. Sigma Vector Filters images are nonstationary in nature due to the presence of 8. EntropyVector Filters edges and fine details, and also the human visual system is 9. Quaternion based Vector Filters nonlinear, nonlinear filters are preferred more than linear filters. A color image can be treated as a mapping for 2D 10. Morphological based vector median filters to 3D [5]. The three color channels exhibit strong spectral 11. Wavelet based median filters correlation. Marginal or component-wise filtering methods 12. Miscellaneous Filters which process each channel independently produce images containing color shifts and other serious artifacts. 2.1 BasicVector Filters In vector filtering techniques the input pixels are treated as a set of vectors and no new colors are introduced. Hence The reduced vector or aggregated ordering technique is theyare able to preserve the correlation among the channel the most common filtering approach. In this method the components [5-8]. Thus an efficient filter aims at aggregated distance of a sample pixel inside a sliding processing of color images with respect to its trichromatic filtering window of length N, usually a finite odd ๐๐ number, is computed as follows: ๐๐ ๐๐ Manuscript received December 5, 2016 Manuscript revised December 20, 2016 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 67
= , , = 1, โฆ , (1) 2.1.3 Extended Vector Median Filter (EXVMF) in which๐๐ ฯ(.) represents the distance or dissimilarity ๐๐ ๐๐=1 ๐๐ ๐๐ function๐ท๐ท โ , ๐๐๏ฟฝ=๐๐( ๐๐,๏ฟฝ ๐๐, ) and๐๐ =( , , ) for three EXVMF combines the vector median operation with an averaging filter. EXVMF of , โฆ is denoted as channels. The sorting of aggregated ๐๐ ๐๐1 ๐๐2 ๐๐3 ๐๐ ๐๐1 ๐๐2 ๐๐3 such that [9, 10] distance ๐๐ , ๐ฅ๐ฅ โฆ๐ฅ๐ฅ, ๐ฅ๐ฅ in ascending๐๐ ๐ฅ๐ฅ ๐ฅ๐ฅorder๐ฅ๐ฅ represents ๐๐ ๐๐ ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ = ๐๐ ๐๐ ๐๐ same ordering of the associated vectors 1 2 ๐๐ , < ( ) ๐ ๐ ( ๐ท๐ท) ๐ท๐ท ๐ท๐ท( ) ( ) ( ) ( ) ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ (5) ๐๐ , otherwise๐๐ ๐๐ (2) ๐ด๐ด๐ด๐ด๐ด๐ด ๐๐=1 ๐ด๐ด๐ด๐ด๐ด๐ด ๐๐ 2 ๐๐=1 ๐๐๐๐๐๐ ๐๐ 2 1 2 ๐๐ 1 2 ๐๐ ๐๐ ๐๐๐๐ โ โฅ ๐๐ โ ๐๐ โฅ โ โฅ ๐๐ โ ๐๐ โฅ ๐ท๐ท โค ๐ท๐ท โค โฏ โค ๐ท๐ท โ ๐๐ โค ๐๐ โค โฏ โค ๐๐ ๏ฟฝ ๐๐๐๐๐๐๐๐ 2.1.1 Vector Median Filter (VMF) where = and is the vector median 1 ๐๐ output. ๐ด๐ดIt๐ด๐ด๐ด๐ด behaves๐๐ =1like๐๐ VMF ๐๐๐๐๐๐near the edges while in In [9] Vector Median filter (VMF), Generalized Vector smooth๐๐ areas it๐๐ behavesโ ๐๐ like ๐๐the Arithmetic Mean Filter Median Filter (GVMF) and Extended Vector Median (AMF). Filter (EVMF) are introduced for processing vector-valued signals having properties similar with median filters 2.1.4 Generalized Vector Median Filter (GVMF) operation such as zero impulse response and good smoothing ability while preserving sharp edges in the The GVMF [11] of vectors , โฆ is vector such signal. They are based on the concept of nonlinear order that { = 1, โฆ , } and for all = 1, โฆ , ๐๐ ๐๐ ๐บ๐บ๐บ๐บ๐บ๐บ๐บ๐บ statistics and derived as maximum likelihood estimates satisfying the condition ๐๐ ๐๐ ๐๐ ๐บ๐บ๐บ๐บ๐บ๐บ๐บ๐บ ๐๐ from exponential distributions. Since vectors which vary ๐๐ โ ๐ฅ๐ฅ โฃ ๐๐ ๐๐ ๐๐ ๐๐ greatly from the data population correspond to the ( ) ( ) (6) maximum aggregated magnitude difference, the VMF ๐๐ ๐๐ ๐๐=1 ๐บ๐บ๐บ๐บ๐บ๐บ๐บ๐บ ๐๐ ๐๐=1 ๐๐ ๐๐ output is the lowest ranked vector with minimum โwhere๐๐ ๐๐( , ) isโ the๐๐ distanceโค โ ๐๐ between๐๐ โ ๐๐ the vectors and . aggregated distance to the input vectors present inside the window. If , โฆ , represent the vectors inside the 2.1.5 Fast๐๐ ๐๐ Modified๐๐ Vector Median Filter (FMVMF)๐๐ ๐๐ filtering window W, the vector median is computed as 1 ๐๐ follows: ๐๐ ๐๐ In [12] a new filter similar with VMF is developed whose a) For each vector element calculate the sum of computational complexity is lower than that of VMF. The distances to all other vectors inside the filtering window, distance associated with central pixel is denoted by ๐๐ using the Minkowski metric (either๐๐ the or norm) and ๐๐+1 add them together to get sum of distances = + ( ( , )๐๐+2 1 2 ( )/ ๐๐+1 ๐ฟ๐ฟ ๐ฟ๐ฟ ๏ฟฝ 2 ๏ฟฝโ1 = (3) ๐๐+1 ๐๐ ๐๐+1 2 ( , ) ๐๐ ๐๐ ๐๐ โ๐ฝ๐ฝ โ๐๐=1 where๐๐ ๐๐ is2 a๐๐ threshold parameter ๐๐ = 1 = 2 where๐๐ ๐๐=1 ๐๐ for๐๐ city block distance and for ๐๐ ๐๐+1 ๐๐ โ ๏ฟฝ๐๐ โ ๐๐ ๏ฟฝ๐พ๐พ ๐๐+1 ๐๐ Euclidean distance. โand๐๐= ๏ฟฝthe2 ๏ฟฝdistance+1 ๐๐ ๐๐ 2 associated๐๐ with๐ฝ๐ฝ the neighbors of is ๐พ๐พ ๐พ๐พ ๐๐+1 b) Find a parameter such that denotes the given as = ( , ) , = minimum . ๐๐ 2 ๐๐๐๐๐๐ 1,2, 3, โฆ , , excluding ( + 1)/2. ๐๐ Then๐๐ for๐๐=1 some๐๐ ๐๐, if c) Corresponding to ๐๐๐๐๐๐, = (๐๐) represents the ๐๐ โ ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ is smaller than ( )/ i.e. = ( , ) < vector median๐๐ . ๐๐ ๐๐๐๐๐๐ ๐๐๐๐๐๐ 1 , ๐๐ ๐๐ . ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ ( )/ then is ๐๐replaced+1 2 by ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ โ๐๐=1 ๐๐ ๐๐ ๐๐ ๐๐๐๐๐๐ ๐๐+1 2.1.2 ฮฑ-trimmed๐๐ Vector Median Filter (ฮฑ-VMF) ๐๐+1 2 ๐๐ ๐๐ ๐๐ 2 ๐๐ 2.1.6 Directional Vector Median Filter (DVMF) In ฮฑ-VMF a trimming operation is incorporated in which (1+ฮฑ) nearest samples to the vector median are given as In this filter, four vector median are applied across the input to an average filter. The output is defined as follows four main directions of the filtering window at [9, 10]: 0ยฐ, 45ยฐ, 90ยฐand135ยฐ to obtain four vector median output = , [0, 1] (4) , , and . In the second stage, the final output is ( ) 1+๐ผ๐ผ 1 generated by applying another vector median on the four The trimming operation enhances in removing the long 1 2 3 4 ๐๐ฮฑVMF โ๐๐=1 1+๐ผ๐ผ ๐๐๐๐ ๐ผ๐ผ โ ๐๐ โ ๐๐filtered๐๐ ๐๐ results.๐๐ Hence the Directional Vector Median tailed or impulsive noise while the averaging filter Filter (DVMF) output is denoted as [13] performs well with Gaussian noise. = ( ) (7) , , and . w๐ท๐ทhere๐ท๐ท๐ท๐ท๐ท๐ท ( ) 1is the vector median of ๐๐DVMF is๐๐ effective in removing impulsive noise while 1 ๐๐ 2 3 4 preserving๐๐ thin lines. ๐๐ ๐๐ ๐๐ ๐๐ 68 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016
2.1.7 Rank Conditioned Vector Median Filter (RCVMF) the current pixel based on the past and future pixel values in the neighborhood of the current pixel by using a block- It incorporates a decision making process in which every by-block autocorrelation function. The difference between pixels in the filtering window is assigned a rank the predicted pixel and the original current pixel is used as depending on the ordered distance. In RCVMF [14] the a measure for impulse detection. output is the vector median when the rank of central pixel The predicted pixel value at central location ( , ) is is larger than a predefined rank of uncorrupted vector computed as pixels in the filtering window. Mathematically it is ( , ) = ( , ) ( , ). ( , ) = ๐๐ ๐๐(12) denoted as , > ๐๐ ๐๐ โ๐๐2 ๐๐ ๐๐ ( )/ ๐๐w๏ฟฝ here๐๐ ๐๐ โ represents๐๐ ๐๐ ๐๐the๐๐ ๐๐vectorโ ๐๐ ๐๐ โobtained๐๐ ๐ฒ๐ฒ ๐๐from the = , otherwise (8) ๐๐๐๐๐๐ ๐๐+1 2 ๐๐ prediction coefficients, is the noncausal region for ๐๐ ๐๐๐๐ ๐๐ ๐๐ ๐๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐๐+1 ๐๐ ๐๐where ( ๏ฟฝ)/ denotes rank of center pixel and is the linear prediction, denotes the matrix of vector pixels ๐๐ 2 ๐๐2 rank of predefined healthy pixel. used for prediction and๐๐ is the order of prediction. ๐๐ ๐๐+1 2 ๐๐๐๐ Then the predictor๐ฒ๐ฒ decides if the current sample is 2.1.8 Rank Conditioning and Threshold Vector Median corrupted or not and is๐๐ replaced by the vector median if Filter (RCTVMF) the predicted error ( , ) exceeds a pre-defined threshold . Hence the output of the noncausal linear prediction In [14], a new improvement in the RCVMF is proposed in based vector filter ๐๐ ๐๐ ๐๐ is denoted by which the distance between the central pixel and ๐๐ , if ( , ) > = ๐๐๐๐๐๐๐๐ predefined healthy pixel is used as additional criteria for ( , ), ๐๐otherwise (13) impulse detection. ๐๐ ๐๐๐๐๐๐๐๐ โฅ ๐๐ ๐๐ ๐๐ โฅ ๐๐ ๐๐๐๐๐๐๐๐๐๐ ๏ฟฝ , > and > 2.1.11 Basic๐๐ ๐๐Vector๐๐ Directional Filter (BVDF) = ๐๐+1 (9) ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ , otherwise ๐๐ ๐๐ ๐๐๐๐ ๐๐ 2 ๐๐ ๐ท๐ท ๐๐ It is a rank ordered filter in which the angle between two ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ w๐๐ here =๏ฟฝ (๐๐+1 , ) denotes the distance between vectors is used as the distance measure. The vectors with ๐๐ 2 ( ) ๐๐+1 atypical directions are regarded as an outlier and filtering central pixel and neighboring๐๐ healthy pixels in which ๐ท๐ท ๐ฅ๐ฅ ๐๐ 2 ๐๐ is done similar with the VMF. The aggregated sum of ( )(1 < < ) is a rank-ordered and healthy vector angles between the vectors is given by pixel. is a pre-determined threshold. If the central pixel = ( , ) , = 1, โฆ , (14) ๐๐ ๐๐ ๐๐ ๐๐ is detected as impulse, it is replaced by the vector median where ๐๐ ๐๐ ๐๐=1 ๐๐ ๐๐ output๐๐. ๐๐ โ ๐ด๐ด ๐๐ ๐๐ ๐๐ . ๐๐ , = cos (15) 2.1.9 Crossing Level Median Mean Filter (CLMMF) โ1 ๐๐๐๐ ๐๐๐๐ w๐ด๐ดhere๏ฟฝ๐๐๐๐ ๐๐๐๐๏ฟฝ , represents๏ฟฝโฅ๐๐๐๐โฅโฅ๐๐๐๐โฅ ๏ฟฝthe angle between the vectors Crossing Level Median Mean Filter [15] combines the and . The angles are ordered which correspond to ๐ด๐ด๏ฟฝ๐๐๐๐ ๐๐๐๐๏ฟฝ idea of the VMF and Arithmetic Mean Filter (AMF) ordering of the input vectors as follows: ๐๐๐๐ ๐๐๐๐ which is based on the vector ordering technique. If ( ) ( ) ( ) โฆ ( ) ( ) ( ) ( ) โฆ denotes the weight of elements of the ordered vectors ( ) ๐๐ ๐๐ 1 โค ๐๐ 2 โค โฏ ๐๐ ๐๐ โค ๐๐ ๐๐ โ ๐๐ 1 โค ๐๐ 2 โค โฏ ๐๐ ๐๐ ( ), ( ), โฆ ( ), the filt๐ก๐กeredโ output is given as follows: ๐ค๐ค The output of BVDF [17]๐๐ is the vector whose angular ๐๐ distance to all otherโค ๐๐ vector in the window is minimum. ๐๐ 1 ๐๐ 2 ๐๐ ๐๐ ๐๐ = ( ). ( ) (10) BVDF preserves chromaticity better than๐๐ the VMF since ๐๐ vectorโs direction corresponds to its chromaticity. BVDF ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ ๐๐=1 ๐๐ ๐๐ where๐๐ โ ๐ค๐ค ๐๐ considers only directional processing and is effective 1 , for = 1 when vector magnitudes are less important than the ( )( ) ๐๐ vectors direction which is not suitable for multichannel ( ) = (11) ๏ฟฝ ๐๐+1 ๐๐+1+ฮณ , for = 2, โฆ , signal processing. โ ( )( ) ๐๐ ๐๐ 1 where๐ค๐ค ฮณ๏ฟฝ represents the parameter of the filter which ๏ฟฝ ๐๐+1 ๐๐+1+ฮณ ๐๐ ๐๐ 2.1.12 Generalized Vector Directional Filter (GVDF) resembles the standard VMF for ฮณ and AMF for ฮณ= 0. The above mentioned filters do not consider both the โ โ unique features of a vector namely direction and 2.1.10 Vector Filters based on Non-Causal (NC) linear magnitude together which may produce erroneous results. prediction technique GVDF [18] is a generalization of BVDF. In the first step, a set of low- rank vectors that is the first terms are A group of switching filters based on noncausal linear selected from the ordered set of vectors based on the prediction is introduced in [16]. NC gives an estimate of aggregated sum of angular distance as oppose๐๐ d to the IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 69
BVDF in which a single vector with the minimum valued weights, then WVMF is the vector such that aggregated angular sum is selected. Next these vectors are [10, 21] { ; = 1, โฆ , } and for all = 1, โฆ , ๐๐๐๐๐๐ given as an input to an additional filter, e.g. alpha-trimmed ๐๐ (17) ๐๐๐๐๐๐ ๐๐ average filter, multistage median filter or morphological ๐๐ ๐๐ โ ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ ๐๐=1 ๐๐ ๐๐๐๐๐๐ ๐๐ ๐พ๐พ ๐๐=1 ๐๐ ๐๐ ๐๐ ๐พ๐พ filters which consider magnitude processing. Hence Itโ can๐ค๐ค beโฅ summarized๐๐ โ ๐๐ โฅas โคfollows:โ ๐ค๐ค sortโฅ ๐๐ theโ pixels๐๐ โฅ inside the GVDF considers both the directional and magnitude filtering window depending on the value of vector median, processing. duplicate each pixel to the number of their corresponding weight and the median value from the ๐๐ 2.1.13 Directional Distance Filter (DDF) new sequence represents the๐๐ weighted vector median. ๐ค๐ค๐๐ An improved filter is achieved by combining VMF and 2.2.2 ฮฑ-Trimmed Weighted Vector Median Filter (ฮฑ- VDF which is known as Directional Distance Filter (DDF) TWVMF) [18,19]. DDF considers the vectorโs direction and magnitude in which both the vectorโs chromaticity and ฮฑ-TWVMF [21] of vectors , , โฆ , having weights intensity are considered. The combined aggregated , , โฆ , is defined as 1 2 ๐ต๐ต measure is defined as follows: ๐๐ ๐๐ ๐๐ 1 2 ๐๐ = . ( , ) (16) ๐ค๐ค ๐ค๐ค =๐ค๐ค where ๐๐ (0, 1) is a parameter1โ๐๐ which๐๐ tunes the๐๐ influence , if < ๐๐ ๐๐=1 ๐๐ ๐๐ ๐พ๐พ ๐๐=1 ๐๐ ๐๐ ๐ผ๐ผโ๐๐๐๐๐๐๐๐ (18) โฆof magnitude๏ฟฝโ โฅ and๐๐ โ angle๐๐ โฅ quantities.๏ฟฝ ๏ฟฝโ ๐ด๐ด ๐๐ ๐๐ ๏ฟฝ ๐๐ ๐๐, ๐๐ otherwise ๐๐ โ ๐๐๐ผ๐ผ โ๐๐=1 ๐ค๐ค๐๐ โฅ ๐๐๐ผ๐ผ โ ๐๐๐๐ โฅ โ๐๐=1 ๐ค๐ค๐๐ โฅ ๐๐๐๐๐๐๐๐ โ ๐๐๐๐ โฅ where๏ฟฝ ๐๐๐๐๐๐ = and = ; having < 2.1.14 Filters based on Hopfield Neural Network and ๐๐ 1 ๐ผ๐ผ is the sum๐๐๐๐โ๐๐ ๐ผ๐ผof ๐๐weighted๐ผ๐ผ from๐๐ vector ๐๐ to all Improved Vector Median Filter ( ) ๐๐ โฃ๐๐๐ผ๐ผโฃ โ ๐๐ ๐๐ ๏ฟฝ๐๐ ๐๐ , = 1, โฆ , . other๐๐โ๐ผ๐ผ vectors๐๐ represents the ๐๐number In this filter, the noise detection in done using a Hopfield ๐๐ ๏ฟฝ ๐๐ ๐๐ , โฆ of elements in ๐๐ and ( ) is the๐ผ๐ผ smallest of . neural network (HNN) and in the second stage, the noisy can have any๐๐ value๐๐ 0, 1, ๐๐โฆโฆโฆ.,โฃ ๐๐ โฃ๐ก๐ก โN-1. pixels are replaced by an improved Vector Median Filter ๐๐๐ผ๐ผ ๐๐ ๐๐ ๐๐ ๐๐1 ๐๐๐๐ first in RGB space and then in HSI space [20]. For the ๐ผ๐ผ2.2.3 Extended Weighted Vector Median Filter improved VMF, the steps are given by (EXWVMF) 1. The vector median is computed in RGB space inside the filtering window. The EXWVMF [21,22] is an extension of WVMF which 2. All the pixels fit for being median inside the is defined as filtering window are collected. = , if < 3. If more than one pixel is fit for being median, ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ ๐๐ , otherwise๐๐ ๐๐ then select that particular pixel which is nearest ๐๐๐๐๐๐๐๐๐๐ โ๐๐=1 ๐ค๐ค๐๐ โฅ ๐๐๐๐๐๐๐๐๐๐ โ ๐๐๐๐ โฅ โ๐๐=1 ๐ค๐ค๐๐ โฅ ๐๐๐๐๐๐๐๐ โ ๐๐๐๐ โฅ (19)๏ฟฝ to the mean of the Hue in HSI space ๐๐๐๐๐๐๐๐ 4. In Step 3 if more than one pixel is qualified, then where = (20) 1 ๐๐ the pixel which is nearest to the mean of EWVMF๐๐ ๐๐๐๐๐๐chooses๐๐ the average๐๐=1 ๐๐ ๐๐as output in the smooth ๐๐ โ๐๐=1 ๐ค๐ค๐๐ โ ๐ค๐ค ๐๐ saturation in HSI space is selected. areas while it chooses weighted vector median (WVM) near edges. 2. 2 Weighted Vector Filters 2.2.4 Rank Order Weighted Vector Median Filter Weighted Vector Filters are extension of Weighted (ROWVMF) Standard Median Filters in which a non-negative weight is assigned to every pixel inside the filtering window In [23] an adaptive noise attenuating and edge enhancing offering more flexibility. filter based on the minimization of aggregated weighted distances among pixels in window is proposed. The 2.2.1 Weighted Vector Median Filter (WVMF) distance between a pixel and all other pixels inside the filtering window is ordered to obtain ( )by assigning a WVMF is a generalization of VMF in which each pixel ๐๐ rank ๐๐ in the filter window is assigned a positive integer ๐๐๐๐ ๐๐ , , โฆ , ( ), ( ), โฆ , ( ) weight. The weight controls the filtering behavior while ๐๐ ๐ฅ๐ฅ๐๐ A weighted sum of distances is computed by utilizing the offering greater flexibility than the median-based filter. If ๐๐๐๐1 ๐๐๐๐2 ๐๐๐๐๐๐ โ ๐๐๐๐ 1 ๐๐๐๐ 2 ๐๐๐๐ ๐๐ , , โฆ , are vectors inside the filtering window and distance ranks denoted as follows: ( ) , , โฆ , are the corresponding nonnegative integer- = . ( ) (21) 1 2 ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ ๐๐=1 ๐๐ ๐๐ ๐ค๐ค1 ๐ค๐ค2 ๐ค๐ค๐๐ ๐ฌ๐ฌ โ ๐๐ ๐๐ ๐๐ 70 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 where ( ) denotes a constantfunction associated with the noisy. Also noise characteristics varies in the image and distance rank . All the weighted aggregated distances hence nonadaptive filters have low performance. Adaptive are sorted๐๐ ๐๐ forming a new order of vectors filters are introduced to handle the difficulty of varying ๐๐ ( ), ( ), โฆ , ๐๐( ) ( ), ( ), โฆ , ( ) ๐ฌ๐ฌ noise characteristics by implementing estimation โ โ โ procedures based on the nature of data on local image The1 output2 of the๐๐ ROWVMF1 2 is the๐๐ vector ( ). Another ๐ฌ๐ฌ ๐ฌ๐ฌ ๐ฌ๐ฌ โ ๐๐ ๐๐ ๐๐ โ statistics [3]. The coefficients of filter kernel change filter called Rank-based Vector Median Filter1 having similar concept is also designed in [24]. ๐๐ values depending on the image structure which is to be smoothed. 2.2.5 Weighted Vector Directional Filters (WVDF) 2.3.1 Adaptive Vector Median filter (AVMF) It employs a nonnegative real weighing coefficient { , , โฆ } corresponding to vector elements In this filter desired features are made invariant to filtering { , , โฆ } with filter output = which operation while the noisy pixels are effected by altering 1 2 ๐๐ minimizes๐ค๐ค ๐ค๐ค ๐ค๐คthe aggregated weighted angular distance with between VMF and the identity operation. This is based on 1 2 ๐๐ ๐๐๐๐๐๐๐๐ ๐๐ other๐๐ ๐๐ vector๐๐ pixels given by [25, 26๐๐] ๐๐ โ ๐๐ the decision rule expressed as follows [31] if , then is impulse = ( , ) (22) ๐๐ else is noise free๐๐ +1 w๐๐here๐๐๐๐๐๐ ( , ) representsโ๐๐=1 ๐๐ the ๐๐angle๐๐ between two vectors. ๐๐๐๐๐๐ โฅ ๐๐๐๐๐๐ ๐๐ 2 ๐๐ ๐๐๐๐๐๐๐๐๐๐โ๐๐๐๐๐๐๐๐ ๐ค๐ค ๐ด๐ด ๐๐ ๐๐ ๐๐+1 Similarly Weighted Directional Distance Filter (WDDF) is where is the vector distance between the central pixel ๐๐ ๐๐ ๐๐ 2 also obtained๐ด๐ด ๐๐ ๐๐ using both the magnitude and angular and the mean of the first vector order statistics ๐๐๐๐๐๐ distance criteria. ๐๐+1, , โฆ , associated with the ordered distances ๐๐( 2) ( ) ( ) ๐๐ ( ), ( ), โฆ , ( ) , for . is a prespecified 2.2.6 Genetic Algorithm based Weighted Vector ๐๐ 1 ๐๐ 2 ๐๐ ๐๐ Directional Filter (GA WVDF) threshold1 2 value.๐๐ Mathematically, is denoted by ๐ฟ๐ฟ ๐ฟ๐ฟ ๐ฟ๐ฟ ๐๐ โค ๐๐ ๐๐๐๐๐๐ = ( ) (24) An optimized WVDF based on Genetic Algorithm is 1 ๐๐๐๐๐๐ ๐๐+1 ๐๐ ๐๐=1 ๐๐ designed in [27] which adapts the filter weights in order to where ฮณ characterizes2 ๐๐ the norm used in Minkowski metric. ๐๐๐๐๐๐ ๏ฟฝ๐๐ โ โ ๐๐ ๏ฟฝ๐พ๐พ match the varying image and noise characteristics. Since If noise is detected the central pixel is replaced by the GA-based methods search the entire solution space, they vector median output. are able to provide a globally optimal (or very close) Another adaptive filter is the Adaptive Basic Vector solution as compared with other optimization techniques. Directional Filter (ABVDF) which is the angular Another filter based on Genetic Programming (GP) is counterpart of AVMF. developed in [28] aiming at the removal of mixed noise. The estimator is based on the global learning capability of 2.3.2 Adaptive based Impulsive Noise Removal Filter GP and measurement of local statistical properties of the healthy pixels present in the surrounding of the corrupted In [32], a new adaptive filtering scheme is proposed. The pixel. impulse is detected based on the difference between the aggregated distance assigned to central pixel and the 2.2.7 Center-Weighted Vector Median Filter (CWVMF) vector median output. The output of the filter is given as follows: In WVMF when the center weight is varied and the others = + (1 ) ( ) (25) remain fixed, a new class of filter is formed called the ๐๐+1 w๐๐๐๐here๐๐๐๐๐๐ ๐๐ is a filter parameter and1 is the VMF output. Center weighted Vector Median Filter (CWVMF) [29,30] ๐๐ ๐ผ๐ผ๐๐ 2 โ ๐ผ๐ผ ๐๐ ( ) is formed. It is defined as The value of is 0 when an impulse1 is present otherwise ๐ผ๐ผ ๐๐ = argmin ( ). (23) it is 1. ๐ผ๐ผ ๐๐ ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ 2 ๐๐+=12, ๐๐for = (๐๐ +๐๐1)/2 2.3.3 Multiclass Support Vector Machine based Adaptive ๐๐with (๐๐ ) = ๐ฅ๐ฅ๐๐โ๐๐ ๏ฟฝโ ๐ค๐ค ๐๐ โฅ ๐๐ โ ๐๐ โฅ๏ฟฝ 1, otherwise Filter (MSVMAF) ๐๐ ๐๐ โ ๐ ๐ ๐๐ ๐๐ where๐ค๐ค only๐๐ the๏ฟฝ central weight ( )/ is varied with smoothing parameter . A new filter called Multiclass Support Vector Machine ๐ค๐ค ๐๐+1 2 based Adaptive Filter (MSVMAF) is developed in [33] for 2.3 Adaptive Vector๐ ๐ Filters reducing high density impulse noise. It takes the advantages of both multiclass Support Vector Machine VMF and its variants result in fixed amount of smoothing (SVM) as well as adaptive vector median filtering. leading to blurring of edges and fine details since they perform filtering operation on all pixels which may not be IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 71
2.3.4 Adaptive Threshold and Color Correction (ATCC) filters are designed in [29,38,39]. They are based on user- Filter defined threshold for detection of impulses. If the central pixel is detected as corrupted, it is replaced by one of the For removing random-valued impulse noise, a new filter output of VMF, BVDF and DDF forming ACWVMF, named Adaptive Threshold and Color Correction (ATCC) ACWBVDF and ACWDDF. The mathematical filter is proposed in [34]. It has an adaptive threshold expressions of the corresponding filters are given below which is computed on the basis of local pixel statistics within the sliding window. , if >
= ๐๐+2 ๐๐+1 ๐๐๐๐๐๐, otherwise๐๐=๐๐ ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ 2.3.5 Robust Switching Vector Filter (RSVF) ๐๐ โ โฅ ๐๐ ๐๐ โ ๐๐ 2 โฅ ๐๐ ๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด ๐๐ ๏ฟฝ ๐๐+1 In this filter, the pixels in the window are ordered (28) ๐๐ 2 according to the distance measure used in VFM, VDF and , if ( ) > DDF. A pixel is detected as corrupted if the cumulative = ๐๐+2 ๐๐+1 ๐ต๐ต๐ต๐ต๐ต๐ต๐ต๐ต, otherwise๐๐=๐๐ ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ ๐๐ โ ๐ด๐ด ๐๐ ๐๐ โ ๐๐ 2 ๐๐ distance ( )/ of the central pixel is larger than the ๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด median cumulative distance of the neighborhood. The (29)๐๐ ๏ฟฝ ๐๐+1 ๐๐+1 2 ๐๐ 2 output of๐๐ the filter is one of the outputs of VMF, VDF and = , if ( ) > DDF as follows: ๐๐๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด ๐๐+2 ๐พ๐พ ๐๐+1 ๐๐+1 1โ๐พ๐พ (30) ๐ท๐ท๐ท๐ท๐ท๐ท, otherwise๐๐=๐๐ ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ๐ถ = ๐๐ โ ๐ด๐ด ๐๐ ๐๐ โ ๐๐ 2 โฅ ๐๐ ๐๐ โ ๐๐ 2 โฅ ๐๐ ( )/ , if ( )/ med . ( , โฆ , ) < 0 ๏ฟฝ ๐๐+1 2 ๐๐๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ (26) w๐๐here ฮป [1, 1] ๐๐+1, 2 otherwise๐๐+1 2 1 ๐๐ ๐๐+1 w๐๐here (.๐๐) is a robustโค univariate๐ผ๐ผ ๐๐ median๐๐ operator and ๏ฟฝ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ 2.3.8. Modifiedโ 2 switchingโ median filter (MSMF) ๐๐ is a filter parameter used for preserving image details ๐๐๐๐๐๐ and smoothing. If = ( , ) the output is It is an extension of the VMF and AVMF consisting of ๐ผ๐ผ ๐๐= denoted by the VMF ( ๐๐ ๐๐=1 ๐พ๐พ ๐๐ ) ๐๐to obtain Robust two-stage noise detector [40]. In the first stage, the Switching Vector ๐๐Medianโ ๐ฟ๐ฟFilter๐๐ ๐๐ (RSVMF). If ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ probably noisy candidates are detected using the AVMF = ( , ), ๐๐ = ๐๐ to represent Robust detection procedure as follows Switching๐๐ Basic Vector Directional Filter (RSBVDF) ๐๐๐๐ โ๐๐=1 ๐ด๐ด ๐๐๐๐ ๐๐๐๐ ๐๐๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐๐๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ , ( ) while for = ( ( , )) ( ( , )) , 1 = ๐๐๐๐๐๐๐๐ ๐๐+1 ๐๐ (31) 1โ๐พ๐พ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ 2 ๐๐=1 ๐๐ = which๐๐ is obtained๐พ๐พ ๐๐by the Directional ๐๐๐๐๐๐๐๐ ๐๐ , ๏ฟฝotherwise๐๐ 2 โ ๐๐ โ ๐๐ ๏ฟฝ โฅ ๐๐๐๐๐๐ ๐๐๐๐ โ๐๐=1 ๐ด๐ด ๐๐๐๐ ๐๐๐๐ โ๐๐=1 ๐ฟ๐ฟ๐๐ ๐๐๐๐ ๐๐๐๐ ๐พ๐พ Distance filter output to represent Robust Switching ๐๐ ๏ฟฝ ๐๐+1 ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ In the second phase for preserving the edge pixels, these Directional๐๐ ๐๐ Distance filter (RSDDF) [35,36]. ๐๐ 2 noise candidates are again judged by four Laplacian 2.3.6 Adaptive Marginal Median filter (AMMF) operators in which the central pixel is convolved with four convolution kernals. For edge detection, the minimum A new modification to Vector Marginal Median Filter difference of the four convolutions denoted by is used (VMMF) is designed in [4] which aims at integrating the and the noisy samples are replaced by the VMF. The noise reduction capability of VMMF as well as preserving output is defined as follows: ๐๐ the vector correlation resulting from VMF. From the = min | = 1, โฆ ,4 (32) ordered aggregated distance used in VMF ๐๐+1 , ๐๐ ( ), ( ), โฆ ( ) ( ), ( ), โฆ ( ) , select a set of ๐๐ ๏ฟฝ๐๐ 2 โ ๐๐ ๐๐ ๏ฟฝ = ๐๐๐๐๐๐, otherwise (33) vectors1 2 constituted๐๐ by1 2vectors๐๐ which are most similar ๐๐๐๐๐๐๐๐ ๐๐ ๐๐ ๐๐ โ ๐๐ ๐๐ ๐๐ ๐๐ ๐๐+1 ๐๐ โฅ ๐๐ to the Vector Median ( ) such that = ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ 2 ๐๐ ๏ฟฝ ๐๐ 2 { , ๐๐ , โฆ } ๐๐ . 2.3.9 Sharpening Vector Median Filter ( ) ( ) ( ) for Then1 the Vector Marginal Median filter is applied to this set๐๐ to achieve high noise๐๐ 1 2 ๐๐ reduction.๐๐ ๐๐ The๐๐ output ๐๐ofโค the๐๐ marginal median filter is In [23] if the function ( ) is a step-like function defined given as follows [37] by 1, for ๐๐ ๐๐ = ( ) = (34) ( ({ , โฆ }) , ({ , โฆ }) , ( ({ , โฆ }))) 0, otherwise ๐ด๐ด๐ด๐ด๐ด๐ด ( ) ( ) ( ) ( ) ( ) ( ) ๐๐ ๐ ๐ ๐ ๐ ๐บ๐บ ๐บ๐บ ๐ต๐ต ๐ต๐ต then a new vector๐๐ โค ๐ผ๐ผ ๐๐filter๐๐๐๐ ๐ผ๐ผ isโค ๐๐obtained known as the (27) 1 ๐๐ 1 ๐๐ 1 ๐๐ ๐๐ ๐๐ ๏ฟฝ ๏ฟฝ๐๐๐๐๐๐ ๐ฅ๐ฅ ๐ฅ๐ฅ ๏ฟฝ ๏ฟฝ๐๐๐๐๐๐ ๐ฅ๐ฅ ๐ฅ๐ฅ ๏ฟฝ ๐๐๐๐๐๐ ๐ฅ๐ฅ ๐ฅ๐ฅ Sharpening Vector Median Filter [41]. 2.3.7 Adaptive Center-Weighted Vector Filter
To provide more flexibility for modification in the size and shape of the window, adaptive center weighted vector 72 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016
2.3.10 Adaptive rank weighted switching filter (ARWSF) Linear Discriminant computed on the aggregated distance and the outliers are replaced by the VMF. A modification ARWSF [42] is a modification of the Rank Order to peer group filter is the Fast Peer Group Filters (FPGF) Weighted Vector Median Filter which incorporates an [47] in which the central pixel is regarded as noise free if adaptive scheme. If the rank weighted distance assigned to pixels are found to be similar. If the noise density is the central pixel is denoted by , then the low, is kept low which reduces the number of distance ๐๐+1 ๐๐+1 calculation.๐๐ In [48] a new Fast Averaging Peer Group difference = ( ) is used for measuring impulse ๐๐ 2 ๐ฌ๐ฌ 2 Filter๐๐ (FAPGF) is designed in which a pixel is considered ๐๐+1 noise corruption. The output1 of ARWSF is given as ๐ฟ๐ฟ ๐ฌ๐ฌ 2 โ ๐ฌ๐ฌ as noisy if the peer group consists of at least two close follows: pixels. If this condition is not satisfied, the central pixel is , if > replaced by the weighted average of the uncorrupted = , otherwise (35) pixels from the neighborhood. In order to improve the ๐๐๐ด๐ด๐ด๐ด๐ด๐ด ๐ฟ๐ฟ ๐๐ where๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด is๐๐ +the1 output of the Arithmetic Mean Filter efficiency and detection, fuzzy metric can be used for ๐๐ ๏ฟฝ ๐๐ 2 which is calculated using the non-corrupted pixels defining the peer group concept. ๐ด๐ด๐ด๐ด๐ด๐ด declared๐๐ by the detector. A two stage filter based on the fuzzy peer group concept is developed in [49] for removing Gaussian and impulse 2.4. Peer Group Vector Filters noise as well as mixed Gaussian impulse noise. It consists of a fuzzy rule โbased switching impulse noise filter Peer Group Filters use the neighborhood of each pixel followed by a fuzzy averaging filter. A fuzzy peer group is while building its peer group which is defined as a set defined as a fuzzy set which considers a peer group as constituted by the central pixel with neighboring pixels support set and membership degree of each peer group which are similar to it [43-45]. member is given by its fuzzy similarity with respect to the central pixel. A Novel Peer Group Filter (NPGF) is 2.4.1 Peer Group Averaging (PGA) Filter proposed in [50] in which the noise detection is done in the CIELab instead RGB color space in order to enhance For an image , the peer group associated with a pixel the noise detection ability. The peer group is formed by comprises of pixels in a predefined ร window centered two different sized windows aiming at reducing the false at , whose intensity๐ผ๐ผ is nearest with . This pixel is then๐๐ detection of non-corrupted pixels. replaced by the intensity of average๐๐ of๐๐ the peer group. This๐๐ concept is referred to as Peer Group๐๐ Averaging (PGA) 2.5. Fuzzy Vector Filters [44]. Fuzzy set concepts are suitable for dealing with ambiguity 2.4.2 Peer Group Vector Filter resulting from inexactness and imprecision in an image such as shape of objects, continuity of line segment and For color images, Peer Group filters [43] use the vector similarity of regions [2]. Also based on the fact that the filter such as VMF. To develop the peer group, first the human vision system is a fuzzy system, many fuzzy based pixels in window are sorted in ascending order according filters have been designed for image enhancement. The to the distance between the central pixel and the weights of Fuzzy filtersare determined from the features neighboring pixels as follows: of datainside the window by applying fuzzy = for = 1,2, โฆ (36) transformation thus utilizing the local correlation [51]. ๐๐+1 This transformation can be modeled as a membership Then๐๐ the peer group๐๐ ๐พ๐พ of the central pixel is computed as ๐๐ โฅ ๐๐ 2 โ ๐๐ โฅ ๐๐ ๐๐ function in accordance to a specific window component.A pixels that rank lowest in the ordered sequence with fuzzy set is defined by the membership function : given by ๐๐ [0,1] that transforms the pixels in image to fuzzy set + 1 ๐๐ ๐น๐น = with degree of membership ranging between 0 and 1.ยต ๐ผ๐ผ โ 2 ๐ผ๐ผ To check๏ฟฝโ๐๐ the๏ฟฝ presence of impulse, the first order ๐๐ 2.5.1 Fuzzy Weighted Average Filter [FWAF] difference of the peer group is calculated = for = 1,2, โฆ The general form of fuzzy based filters is a fuzzy The central pixel is declared as noisy if any one of these ๐๐ ๐๐+1 ๐๐ weighted average [52,53] of the pixel values inside the ๐ฟ๐ฟdifferences๐๐ โ is๐๐ larger๐๐ than a ๐๐pre-specified threshold and filtering window. The output is estimated as the centroid replaced by VMF. given below In [46], a similar peer group switching filter is proposed ( ) = = (37) which utilizes the statistical properties of the sorted ๐๐ ( ) ๐๐ โ๐๐=1 ๐๐ ยต๐๐ ๐๐๐๐ sequence of the aggregated distance of pixels inside ๐น๐น๐น๐น๐น๐น๐น๐น ( ๐๐=)1 =๐๐ ๐๐ ๐๐ ๐๐where โ ๐ค๐ค ๐๐ denotesโ๐๐=1 ๐๐ ยต ๐๐ an adaptive function filtering window. Noise detection is based on the Fisherโs determined from ๐๐the local context with membership ๐๐ ยต๐๐ ยต๐๐ IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 73 function of the pixel and ฮป is a parameter such that One kind of these filters proposed in [15,55] can be ฮป [0, ). This filter should satisfy the following two described as follows: ๐๐ ๐๐ constraintsยต ๐ฅ๐ฅ Perform a pre-processing activity for removing impulse โ (i) โ Each weight is a positive number i.e. 0 noise from the set of pixels in the filtering window by and forming a new set = { ; = 1, โฆ , } . This is ๐๐ (ii) The sum of all the weights is equal๐ค๐ค โฅ to obtained by replacing โฒthe minimumโฒ and maximum๐๐ ๐๐ unity = 1. luminance values by the๐๐ median๐๐ ๐๐pixel value๐๐ . The ๐๐ output of the filter is given by ๐๐=1 ๐๐ ๐๐๐๐๐๐ 2.5.2 Fuzzy Stack Fโilter๐ค๐ค ( ) ๐๐ = (41) ๐๐ ( โฒ )โฒ โ๐๐=1 ยต๐ฑ๐ฑ ๐ฅ๐ฅ๐๐๐๐ ๐๐๐๐ In [54] Fuzzy Stack Filters are proposed to extend the w๐น๐นhere๐น๐น๐น๐น ๐๐ = โฒ and is the membership ๐๐ โ๐๐=1 ยต๐ฑ๐ฑ ๐ฅ๐ฅ๐๐๐๐ smoothing characteristics of stack filters which is based on function whichโฒ โฒ describes a ฮ -type (i.e. a bell-shaped) ๐๐ ๐๐ ๐๐๐๐๐๐ ๐ฑ๐ฑ the application of fuzzy positive Boolean function. fuzzy set๐ฅ๐ฅ๐๐ aiming๐๐ โ ๐๐at removingยต the pixels with large luminance values. 2.5.3 Fuzzy Vector Median Filter (FVMF) 2.5.7 Adaptive Nearest-Neighbor Filter (ANNF) In Fuzzy Vector Median Filter the dissimilarity distance measure used is the Minkowski metric which is fed as Adaptive nearest-neighbor filter (ANNF) is based on the an input to the membership function for determining the ๐พ๐พ nearest neighbor rule in which the fuzzy weights are fuzzy weights. The membership ๐ฟ๐ฟfunction is the calculated as follows [56]: exponential (Gaussian-like) form [52, 53]: = ( ) ( ) for = 1, 2, โฆ . (42) ( ) ( ) ( ) = exp [ ] (38) ๐๐ ๐๐ โ๐๐ ๐๐ ๐๐ ๐๐ ๐ฟ๐ฟ๐พ๐พ ๐๐ ๐ค๐คwhere๐๐ ๐๐( โ)๐๐ is1 the๐๐ maximum๐๐ angular distance and ( ) w๐๐here denotes a positive constant and is a distance represents the minimum angular distance associated with ยต โ ๐๐ ๐๐ 1 threshold which control the amount of fuzziness in the the center๐๐ -most pixel inside the filtering window. ๐๐ weights.๐๐ ๐๐ 2.5.8 Adaptive Nearest-Neighbor Multichannel Filter 2.5.4 Fuzzy Vector Directional Filter (FVDF) (ANNMF)
In Fuzzy directional filter the vector angle criterion An improvement in the ANNF is the Adaptive nearest- (angular distance) is used as the distance function and has neighbor multichannel filter (ANNMF) which combines asigmoidal membership function. The fuzzy weight vector directional with vector magnitude filtering. The associated with the vector is given by [52, 53] distance measure for noisy vector is given by the = (39) following formula [57] ( ( )) ๐๐ ๐๐ ๐๐ ๐๐ = (1 ( , )) (43) ๐๐ wh๐๐ ere is the ๐๐angular distance measure. ยต 1+exp โบ๐๐ ๐๐ ๐๐ , ๐๐=1= ๐๐ 1๐๐ (44) ๐๐ ๐๐ โ โ ๐๐๐ก๐ก ๐๐ ๐๐ ( , ) 2.5.5 Fuzzyโบ Ordered Vector Filter (FOVF) ๐๐๐๐๐๐๐๐ โฃโฅ๐๐๐๐โฅโโฅ๐๐๐๐โฅโฃ ๐๐The๏ฟฝ๐๐ ๐๐first๐๐๐๐๏ฟฝ part๏ฟฝ โฃof๐๐๐๐โฃ โฃ๐๐the๐๐โฃ๏ฟฝ equation๏ฟฝ โ max representsโฃ๐๐๐๐โฃ โฃ๐๐๐๐โฃ ๏ฟฝ the vector angular Fuzzy Ordered Vector Filters use only a part of the criteria and second part is the normalized magnitude vector-valued pixels which are ordered according to their difference. According to this equation the directional corresponding fuzzy membership strengths. It is given as information is used when the vectors have same length follows: while the magnitude difference part is used when the
= ( ) ( ) (40) vectors have direction. 1 ๐๐ where = ( ) ๐น๐น๐น๐น๐น๐น๐น๐น๐น๐น ๐๐ โ๐๐=1 ๐๐ ๐๐ 2.5.9Adaptive Hybrid Multichannel Filter (AHMF) ๐๐ ๐๐ ๐ค๐คth ๐๐ ( ) denotes ๐๐the=1 i ๐๐ ordered fuzzy membership function ๐๐ โ ๐ค๐ค To achieve three objectives such as noise attenuation, such๐๐ that ( ) ( ) ( ) with ( ) being the fuzzy๐ค๐ค coefficient having the largest membership value chromaticity retention and edges or detail preservation a ๐๐ ๐๐โ1 1 1 [52].These๐ค๐ค filtersโค ๐ค๐ค resembleโค โฏ fuzzyโค ๐ค๐ค generalization๐ค๐ค of ฮฑ- new filter named Adaptive Hybrid Multichannel Filter trimmed filters. (AHMF) is introduced in [58]. The structure of AHMF consists of three parts: a Hybrid Multichannel (HM) filter, 2.5.6 Fuzzy Hybrid Filter (FHF) a fuzzy ruled-based system and a learning algorithm.HM comprises of four components: VMF, BVDF, Identity Hybrid filters combine a nonlinear filter used for Filter (IF) and a summation combinatory. suppression of noise with a fuzzy weighted linear filter. 74 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016
2.5.10 Fuzzy Impulse Detection and Reduction Method 2.6.2 Structure-Adaptive Hybrid Vector Filter (SAHVF) (FIDRM) In [61], another hybrid filter is designed. At each pixel For images corrupted with mixture of impulse noise and location, noise adaptive preprocessing and modified quad other types of noise, Fuzzy Impulse Detection and tree decomposition techniques are used for classifying the Reduction Method [59]is developed based on the concept corrupted image into several signal of fuzzy gradient values which constructs a fuzzy set activity categories. Then according to the structure impulse noise. This fuzzy set is denoted by a fuzzy classification, window adaptive hybrid filter is designed. membership function which is used by filtering method The hybrid filter consists of three sub filters as follows usually a fuzzy averaging of neighboring pixels. It results a) Modified Peer Group (MPG) which reduces an output image quasi without or with very little impulse image degradation in high-activity regions noise such that other filters can be applied afterwards. b) Adaptive Nearest Neighbor Filter (ANNF) which removes noise by preserving edge structures in 2.6. Hybrid Vector Filters medium-activity regions c) Structure Weighted Average Filter (SWAV) Hybrid Filters are combination of sub filters which belong which smoothes small distortions in low-activity to different types giving an output being a linear or non- regions linear combination of the vector samples.
2.6.1 Vector Median Rational Hybrid filter (VMRHF)
Vector Median Rational Hybrid filters [60] are multispectral image processing filters based on rational functions. VMRHF comprises of three sub filters (two vector median filters and one centre weighted vector median filter) with a vector rational operation. The output of VMRHF results from the vector rational function considering three input sub functions ๐๐๐๐๐๐๐๐๐๐ { , , ๐๐ } where is fixed as center weighted Fig. 2 Schematic diagram of Structure-Adaptive Hybrid Vector Filter vector1 sub2 filter3 and is given2 as ษธ ษธ ษธ ษธ ( ) 2.7 SigmaVector Filters = ( ) + (45) 3 ( ) ( ) โ๐๐=1 ๐๐๐๐ษธ๐๐ ๐๐ ๐๐with๐๐๐๐๐๐๐๐ ๐๐ = [ษธ2, ๐๐ , โ]+๐๐โฅ ษธrepresents1 ๐๐ โษธ3 ๐๐ โฅ2the constant vector coefficient used to weight๐๐ the output of the three sub 1 2 3 filters.โบ andโบ โบare โบpositive constants, the former ensures numerical stability whereas the nonlinearity is regulated by the latter.โ ๐๐
Fig. 3 Concept of sigma filtering in which radius indicates the variance or variance multiplied by a tuning parameter.
Sigma filters combine the order statistics theory and standard deviation or approximation of multivariate variance. In the figure, the radius of the circles represents the approximation of the variance and if the central pixel Fig. 1 Structure of VMRHF lies inside the circle, it is healthy otherwise it is regarded as an outlier. IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 75
2.7.1 Standard Sigma Filters ( ), for ( )/ = , otherwise (51) Sigma filters are adaptive switching filters which are ๐๐ 1 ๐บ๐บ ๐๐+1 2 โฅ ๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐ ๐๐+1 ( ) based on the concept of standard deviation used for gray- w๐๐here ๏ฟฝ๐๐= 2 ( ) + with = in which ( ) is ๐บ๐บ 1 scale images described in [62, 63]. Mathematically, the the smallest hybrid measureconsidering both magnitude 1 ๐พ๐พ๐พ๐พ ๐พ๐พ๐พ๐พ ๐๐โ1 1 standard deviation is given by and angular๐๐๐๐๐๐ distance๐บ๐บ ๐๐ ๐๐and repre๐๐ sents the approximated๐บ๐บ = ( ( ) ) (46) variance. ๐๐๐ด๐ด 1 ๐๐ 2 ๐๐ ๐๐where๏ฟฝ ๐๐ โrepresents๐๐=1 ๐๐ โ ๐๐๏ฟฝthe mean of the pixels inside the 2.7.3 Adaptive Vector Sigma Filters sliding window. The output of the sigma filter is described as ๐๐๏ฟฝ In Adaptive Vector Sigma Filters [67] the threshold is ( , , โฆ , ), determined adaptively which employ the approximations
= ๐๐+1 of the multivariate variance based on the sample mean or 1 , 2 ๐๐ otherwise ๐๐ ๐๐ ๐๐ ๐๐ ๐๐๐๐ โฃ ๐๐ 2 โ ๐๐๏ฟฝ โฃ ๐๐ โฅ ๐๐๐๐ the lowest-ranked vector. ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐๐(47) ๏ฟฝ ๐๐+1 (i) Design based on the sample mean ๐๐ 2 where (. ) is the smoothing function usually a median filter (a) Adaptive Sigma Vector Median Filter (ASVMF-mean) and is the smoothing factor. The output of ASVMF-mean is given by , if
2.7.2๐๐ Vector Sigma Filters = ๐๐+1 (52) ๐๐๐๐๐๐, ๐พ๐พ ๐พ๐พ ๐๐ โฅ ๐๐ 2 โ ๐๐๏ฟฝ โฅ โฅ ๐๐ ๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ดโ๐๐๐๐๐๐๐๐ The concept of standard sigma filter can be extended to w๐๐here is the mean๏ฟฝ ๐๐+ 1of the samples inside the filtering ๐๐ 2 ๐๐๐๐โ๐๐๐๐๐๐๐๐๐๐๐๐ multichannel images. For calculating the standard window and the variance for chosen norm is given by deviation, the Minkowski metric, the angular distance or = ๏ฟฝ๐๐ (( ) )2 (53) the combination of both can be used. If the central pixel is ๐๐ ๐พ๐พ (b)2 Adaptive1 ๐๐ Sigma Basic2 Vector Directional Filter detected as noisy, it is replaced by one of the output of ๐๐๐พ๐พ ๐๐ โ๐๐=1 โฅ ๐๐๐๐ โ ๐๐๏ฟฝ โฅ ๐พ๐พ VMF, BVDF or DDF otherwise it is left unchanged. (ASBVDF-mean) These filters require a tuning parameter for determining The output of ASBVDF-mean is given by , if ( , ) the switching threshold. These filters are described below = ๐๐+1 (54) [64-66] ๐๐ ๐ต๐ต๐ต๐ต๐ต๐ต๐ต๐ต, otherwise ๐ด๐ด ๐๐ ๐ด๐ด ๐๐ 2 ๐๐๏ฟฝ โฅ ๐๐ (a)Sigma Vector Median Filter (SVMF) ๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ดโ๐๐๐๐๐๐๐๐ w๐๐ith angular definition๏ฟฝ ๐๐+1 of multichannel variance The output of sigma filter is described by ๐๐ 2 defined by 2 ( ), for ( )/ ๐ด๐ด = (48) = ( , ) (55) ๐๐ 1 , otherwise๐๐+1 2 ๐๐ ๐ฟ๐ฟ โฅ ๐๐๐๐๐๐ (c)2 Adaptive1 ๐๐ Sigma2 Directional Distance Filter (ASDDF- ๐๐where๐๐๐๐๐๐๐๐ ๏ฟฝ ๐๐ +is1 a threshold determined from the ๐ด๐ด ๐๐ โ๐๐=1 ๐๐ ๏ฟฝ ๐๐ 2 mean)๐๐ ๐ด๐ด ๐๐ ๐๐ approximation of the multivariate variance of the The output of the ASDDF-mean is given by ๐๐๐๐๐๐ vectors present in the window. = = ( ๐พ๐พ) + = ( ) (49) ๐๐ ๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด,โ๐๐๏ฟฝ ( ( , )) . ( ) ๐๐โ1+๐๐ ๐๐ with ๐๐+1 ๐๐ 1โ๐๐ 1 ๐พ๐พ ๐๐โ1 1 ๐ท๐ท๐ท๐ท๐ท๐ท, otherwise ๐๐ ๐พ๐พ ๐๐๐๐ ๐๐๐๐๐๐ ๐ฟ๐ฟ( ) ๐๐๐๐ ๐ฟ๐ฟ ๐๐ ๐๐๐๐ ๐ด๐ด ๐๐ 2 ๐๐๏ฟฝ โฅ ๐๐ โ ๐๐๏ฟฝ โฅ โฅ ๐๐ = , ( ) is the aggregated distance associated with ๐ฟ๐ฟ 1 ๏ฟฝ(56)๐๐ +1 the vector median ( ) and represents the tuning ๐๐ 2 ๐๐๐พ๐พ ๐๐โ1 ๐ฟ๐ฟ 1 with variance is the combination of and given parameter. ๐๐ 1 ๐๐ by 2 2 2 (b) Sigma Basic Vector Directional Filter (SBVDF) ๐๐๐๐๐๐ ๐๐๐พ๐พ ๐๐๐ด๐ด The output of SBVDF is given by = ( ) ( ) , for (ii)2 Design2 based1โ๐๐ on2 ๐๐ the lowest-ranked vector ( ) ( )/ ๐๐๐๐๐๐ ๐๐๐พ๐พ ๐๐๐ด๐ด = , otherwise (50) In this family, the lowest-ranked vector is used in ๐๐ 1 ๐ผ๐ผ ๐๐+1 2 โฅ ๐๐๐๐๐๐ calculation of the variance. ๐๐๐๐๐๐๐๐๐๐ ๐๐+1 ( ) w๐๐here ๏ฟฝ=๐๐ 2( ) + with = in which ( ) is (a) Adaptive Sigma Vector Median Filter (ASVMF-rank) ๐ผ๐ผ 1 the smallest aggregated angular distance and The output of ASVMF-rank is given by 1 ๐ด๐ด ๐ด๐ด ๐๐โ1 1 , if represents๐๐๐๐๐๐ the๐ผ๐ผ approximated๐๐๐๐ variance๐๐ calculated using๐ผ๐ผ the ( ) ๐ด๐ด = ๐๐+1 (57) angular distance. ๐๐ ๐๐๐๐๐๐, 1 ๐พ๐พ ๐พ๐พ ๐๐ โฅ ๐๐ 2 โ ๐๐ โฅ โฅ ๐๐ (c)Sigma Directional Distance Filter (SDDF) ๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ดโ๐๐๐๐๐๐๐๐ w๐๐here is the๏ฟฝ lowest๐๐+1 -ranked vector inside the filtering The output of SDDF is given by ( ) ๐๐ 2 ๐๐๐๐โ๐๐๐๐๐๐๐๐๐๐๐๐ window 1and the variance for chosen norm is given by ๐๐ 2 ๐๐ ๐พ๐พ 76 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016
, = (( ( ) ) ) (58) ( ) ( )/ ( )/ 1 = , (69) (b)2 Adaptive๐๐ Sigma Basic Vector2 Directional Filter ๐๐ ๐๐+1 2 ๐๐+1 2 ๐๐๐พ๐พ ๐๐โ1 โ๐๐=1 โฅ ๐๐๐๐ โ ๐๐ 1 โฅ ๐พ๐พ ๐๐ ๐๐๐๐ ๐๐ โฅ ๐ฝ๐ฝ (ASBVDF-rank) w๐ธ๐ธhere๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ ๐๐+1 is the adaptive threshold of the central ๐๐ ( ๏ฟฝ๐๐ )/2 ๐๐๐๐โ๐๐๐๐๐๐๐๐๐๐๐๐ The output of ASBVDF-rank is given by pixel and is the vector median output. Similarly other , ( , ) ๐๐+(1 ) 2 ( ) entropy๐ฝ๐ฝ vector filters such as Entropy Basic Vector = ๐๐+1 (59) 1 ๐ต๐ต๐ต๐ต๐ต๐ต๐ต๐ต, otherwise1 ๐ด๐ด Directional๐ฅ๐ฅ Filter (EBVDF) and Entropy Directional ๐๐ ๐๐๐๐ ๐ด๐ด ๐๐ 2 ๐๐ โฅ ๐๐ ๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ดโ๐๐๐๐๐๐๐๐ Distance Filter (EDDF) are also developed according to w๐๐ith angular definition๏ฟฝ ๐๐+1 of multichannel variance ๐๐ 2 their corresponding distance and angular measure [70]. defined by 2 ๐๐๐ด๐ด = ( , ( )) (60) 2.9 Quaternion based Vector Filters 2 1 ๐๐ 2 (c)๐ด๐ด Adaptive๐๐= Sigma1 ๐๐Directional1 Distance Filter (ASDDF- rank)๐๐ ๐๐โ1 โ ๐ด๐ด ๐๐ ๐๐ Another approach for impulsive noise removal is based on The output of the ASDDF-rank is given by the quaternion theory for improving the evaluation of = color dissimilarity. A quaternion number is a four dimensional number that consists of a real part and three , ( ( , ( ))) . ( ( ) ) ๐๐๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ด๐ดโ๐๐๐๐๐๐๐๐ imaginary parts , and [71-73]. In the hyper๐๐ complex ๐๐+1 ๐๐ 1โ๐๐ ๐ท๐ท๐ท๐ท๐ท๐ท, otherwise 01 ๐๐ 1 ๐พ๐พ ๐๐๐๐ ๐๐ ๐๐ ๐๐๐๐ ๐ด๐ด ๐๐ 2 ๐๐ โฅ ๐๐ โ ๐๐ โฅ โฅ ๐๐ form, it is represented as ๏ฟฝ ๐๐+1 = + + ๐๐+๐๐ ๐๐ (70) ๐๐ 2 โฅ (61) with variance is the combination of and given where , and are the complex operators, which obey the following๐๐ ๐๐ ๐๐ rules๐ค๐คฬ ๐๐ ๐ฅ๐ฅฬ ๐๐๐๐๏ฟฝ by 2 2 2 ๐๐๐๐ ๐พ๐พ ๐ด๐ด = ๐ค๐คฬ =ศทฬ =๐๐๏ฟฝ 1 = ( ) ๐๐( ) (62) ๐๐ ๐๐ 2 2 2 2 2 1โ๐๐ 2 ๐๐ = , = , = ๐๐๐๐ ๐พ๐พ ๐ด๐ด ๐ค๐คฬ ๐ฅ๐ฅฬ ๐๐๏ฟฝ โ ๐๐2.8 Entropy๐๐ Vector๐๐ Filters = , = , = Multiplication๐ค๐คฬ๐ฅ๐ฅฬ ๐๐๏ฟฝ ๐ฅ๐ฅฬ๐๐๏ฟฝ ๐ค๐คฬ ๐๐๏ฟฝof๐ค๐คฬ the๐ฅ๐ฅฬ quaternion is not commutative. A ๏ฟฝ ๏ฟฝ Entropy Vector Filters are adaptive multichannel filters pure๐ฅ๐ฅฬ๐ค๐คฬ โquaternion๐๐ ๐๐๐ฅ๐ฅฬ โ has๐ค๐คฬ ๐ค๐ค๐๐ฬ =โ0๐ฅ๐ฅ.ฬ The modulus and conjugate of based on the local contrast entropy introduced by a quaternion are Beghdadi and Khellaf in [68] for gray scale image. Each | | = + + ๐๐+ (71) pixel inside the filtering window is associated with its = 2 2 2 2 (72) contrast defined by ๐๐ โ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ A unit quaternion has unit modulus. A quaternion in the |๐ฅ๐ฅ | ๐๐๏ฟฝ ๐๐ โ ๐๐๐ค๐คฬ โ ๐๐๐ฅ๐ฅฬ โ ๐๐๐๐๏ฟฝ = ๐๐ = (63) polar form is represented as ๐ฅ๐ฅ๐๐โ๐ถ๐ถ๐ฅ๐ฅฬ ๐ฅ๐ฅ๐๐ | | with is the gradient and represents the mean of the = ๐๐ ๐ฅ๐ฅฬ ๐ฅ๐ฅฬ ( ) input๐ถ๐ถ { , , โฆ }. The local contrast probability is = ๐๐๐๐+ (73) ๐๐ ๐๐ ๐๐ ๐๐ given๐ฅ๐ฅ by ๐ฅ๐ฅฬ where is a unit pure quaternion and = + + / 3; ๐ฅ๐ฅ1 ๐ฅ๐ฅ2 ๐ฅ๐ฅ๐๐ and๐๐๐๐๐๐๐๐ are๐๐ ๐๐๐๐๐๐๐๐referred to as the eigenaxis and eigenangle = (64) ๏ฟฝ ๐ฅ๐ฅ๐๐ respectively.ยต ยต ๏ฟฝ ๐ค๐คฬ ๐ฅ๐ฅฬ ๐๐๏ฟฝ โ ๐๐ ๐๐ ๐๐Any โpixel๐๐=1 ๐ฅ๐ฅ ๐๐ is considered as noisy if the local contrast An๐๐ RGB๐๐ color pixel can be represented in quaternion form probability is too high. In case of multichannel image, the as local contrast probability is given by = + + (74) (| |) where , and denote the pixel values of red, green = 1 , (65) 1 1 1 1 ๏ฟฝ ๐๐ ๐พ๐พ ๐พ๐พ ๐๐and blue๐๐ ๐ค๐คฬ channels๐๐ ๐ฅ๐ฅฬ ๐๐ respectively.Any๐๐ unit quaternion U can ๏ฟฝโ๐๐=1 ๐ฅ๐ฅ๐๐๐๐| โ๐ฅ๐ฅฬ ๐๐ |๏ฟฝ 1 1 1 ๐๐ 1 be represented๐๐ ๐๐ ๐๐as = | | . The quaternion unit ๐๐ ๐๐ ๐พ๐พ ๐พ๐พ where๐๐ represents๐๐๐๐ ๐๐ the component of the mean. โ๐๐=1๏ฟฝโ๐๐=1 ๐ฅ๐ฅ โ๐ฅ๐ฅฬ ๏ฟฝ transform of a color pixel can๐๐๐๐ be expressed as follows Noisy pixels contribute heavily๐ก๐กโ to the entropy defined by ๐๐ [74] ๐๐ ๐๐ ๐๐ = ๐๐ ๐๐ 1 (66) = = [ + ๐๐ ] + + [ It is assumed that each sample is associated with an ๐๐ ๐๐ ๐๐ ๐ป๐ป โ๐๐ ๐๐๐๐๐๐๐๐ ] =1 + + 12 + 1 1 + + adaptive threshold defined as the rate of change of local ๐๐ ๐๐๐๐ ๐๐๏ฟฝ ๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐๐ ๏ฟฝ๐๐ ๐ค๐คฬ ๐๐ ๐ฅ๐ฅฬ ๐๐ ๐๐๏ฟฝ๏ฟฝ ๐๐๐๐๐๐๐๐ โ contrast entropy and the overall entropy expressed as 2 ๐๐ +1 [( 1 1 ๏ฟฝ) + ( 3) 1 + 1 ( ๐ฝ๐ฝ ๐๐๐๐๐๐๐๐๐๐ ๏ฟฝ๐๐ ๐ค๐คฬ ๐๐ ๐ฅ๐ฅฬ ๐๐ ๐๐๏ฟฝ๐๐๐๐๐๐ ๐๐ โ ๐๐๏ฟฝ๐๐ ๐ค๐คฬ ๐๐ ๐ฅ๐ฅฬ = (67) 2 1 ๐๐ 1) 2 1 1 1 1 1 โ๐๐๐๐๐๐๐๐๐๐๐๐๐๐ ๐ป๐ป ๐ป๐ป ๐๐ ๐๐๏ฟฝ๏ฟฝ๐ ๐ ๐ ๐ ๐ ๐ ๐๐ โ3 ๐๐ โ ๐๐ ๐ค๐คฬ ๐๐ โ ๐๐ ๐ฅ๐ฅฬ ๐๐ โ The๐๐ output๐๐ of the entropy vector median filter (EVMF) = + + (75) ๐ฝ๐ฝ โ โ๐๐=1 ๐๐๐๐๐๐๐๐๐๐๐๐๐๐ 1 ๏ฟฝ [69] is given by ๐๐where๐๐๏ฟฝ๐ ๐ ๐ ๐ ๐ ๐ ๐๐ = + + 2 , = , if ๐๐๐ ๐ ๐ ๐ ๐ ๐ ๐๐๐ผ๐ผ ๐๐โ ( ) ( )/ ( )/ ( + + ) , = (68) ๐๐๐ ๐ ๐ ๐ ๐ ๐ ๏ฟฝ๐๐1๐ค๐คฬ๐๐ ๐๐1๐ฅ๐ฅฬ๐๐ ๐๐1๐๐๏ฟฝ๐๐๏ฟฝ๐๐๐๐๐๐ ๐๐ ๐๐๐ผ๐ผ 1 , otherwise๐๐+1 2 ๐๐+1 2 2 2 ๐๐ ๐๐ โฅ ๐ฝ๐ฝ 1 1 1 ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ๐ธ ๐๐+1 โ3 ๐๐ ๐๐ ๐ค๐คฬ ๐๐ ๐ฅ๐ฅฬ ๐๐ ๐๐๏ฟฝ ๐ ๐ ๐ ๐ ๐ ๐ ๐๐ ๐๐ ๏ฟฝ๐๐ 2 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 77
= [( ) + ( ) + (g r )k 2 ; for removing mixture noise. In [80] a new two-stage filter 1 is proposed which incorporates the peer group concept represents the intensity of the color image; is the ๐๐โ โ3 ๐๐1 โ ๐๐1 ๐ค๐คฬ ๐๐1 โ ๐๐1 ๐ฅ๐ฅฬ 1 โ 1 ๏ฟฝ๏ฟฝ๐ ๐ ๐ ๐ ๐ ๐ ๐๐ along with the quaternion based distance measure for projection of the tristimuli in Maxwell triangle rotated 90 , ๐๐ ๐ผ๐ผ ๐๐โ impulse detection. The probably noisy pixels are replaced and it represents the chromaticity [3]. When = , =0 by Weighted Vector Median Filter in the second stage. ๐๐ | = 1 2 + 1 6 + + and = 0, = The concept of quaternion is extended in video sequences ๐๐ 4 ๐๐ ๐๐ for removal of random-valued impulse noise in [81]. The (1 3)( + + ) + + = 1๐ ๐ ๐ ๐ ๐ ๐ 3 ( ๐ผ๐ผ ๐๐ ๐๐=4 ๏ฟฝ โโ ๏ฟฝ ๏ฟฝ โโ ๏ฟฝ๏ฟฝ๐ค๐คฬ ๐ฅ๐ฅaฬ nd๐๐๏ฟฝ๏ฟฝ ๐๐ [ ๐๐ luminance and chromaticity difference are represented in ) + ( ) + ( ) ] .Similarly, can be quaternion form which are combined together for โ ๐๐1 ๐๐1 ๐๐1 ๏ฟฝ๐ค๐คฬ ๐ฅ๐ฅฬ ๐๐๏ฟฝ๏ฟฝ ๐๐โ ๏ฟฝ โโ ๏ฟฝ ๐๐1 โ defined as follows measuring color difference between color samples. Based 1 1 1 1 1 ๏ฟฝ ๏ฟฝ ๐๐ =๐ค๐คฬ ๐๐ โ= ๐๐( ๐ฅ๐ฅฬ + ๐๐ +โ ๐๐ )๐๐ + + [๐๐( on this color difference, the color vectors along the 1 1 ) + ( ) + ( ) ] = (76) horizontal, vertical and diagonal directions in current ๐๐๏ฟฝ ๐๐๏ฟฝ๐๐1๐๐ 3 ๐๐1 ๐๐1 ๐๐1 ๏ฟฝ๐ค๐คฬ ๐ฅ๐ฅฬ ๐๐๏ฟฝ๏ฟฝ โ โ3 ๐๐1 โ frame and adjacent frames on motion trajectory are utilize = ( + 1 and 1 1can be1 expressed1 ๐ผ๐ผ asโ for the detection of presence of impulse. For noisy pixels, ๐๐ ๐ค๐คฬ ๐๐ โ ๐๐ ๐ฅ๐ฅฬ ๐๐ โ ๐๐ ๐๐๏ฟฝ ๐๐ โ ๐๐ 1 ๐ผ๐ผ )and โ= ( ). ๐ผ๐ผ 1 a 3-D weighted vector median operation is carried out ๐๐ ๐๐ ๐๐ 2 ๐๐๐๐ ๐๐๏ฟฝ The color pixel1 difference between two pixels and while the other pixels are left intact. Another filter for ๐๐๏ฟฝ๐๐1๐๐ ๐๐โ 2 ๐๐๐๐1๐๐๏ฟฝ โ ๐๐๏ฟฝ๐๐1๐๐ removal of random-valued impulse noise from video can be written as the sum of the intensity and chromaticity ๐๐ ๐๐ sequences is also designed in [82]. differences as follows: ๐๐ ๐๐ , = , + , (76) 2.10 Morphology based filters where , = (1 2)|( + ) ( + ๐๐๏ฟฝ๐๐๐๐ ๐๐๐๐๏ฟฝ ๐๐๐ผ๐ผ๏ฟฝ๐๐๐๐ ๐๐๐๐๏ฟฝ ๐๐โ๏ฟฝ๐๐๐๐ ๐๐๐๐๏ฟฝ )| is the color intensity difference Morphological filters (MF) are designed by parallel or ๐๐๐ผ๐ผ๏ฟฝ๐๐๐๐ ๐๐๐๐๏ฟฝ โ ๐๐๐๐๐๐๐๐๏ฟฝ ๐๐๏ฟฝ๐๐๐๐๐๐ โ ๐๐๐๐๐๐๐๐๏ฟฝ and , = (1 2)|( ) ( sequential combination of the basic fundamental ๐๐๏ฟฝ๐๐๐๐๐๐ )|is the color chromaticity difference.The intensity morphological operations such as erosion, dilation, โ ๐๐ ๐๐ ๐๐ ๏ฟฝ ๏ฟฝ ๐๐ ๐๐ ๏ฟฝ opening and closing [83-85]. The common structures of difference๐๐ ๏ฟฝ ๐๐and๐๐ ๏ฟฝ chromaticityโ difference๐๐๐๐ ๐๐ โ ๐๐ ๐๐approach๐๐ โ ๐๐๐๐ zero๐๐ โ ๐๐ MF are defined as ๐๐when๏ฟฝ๐๐ ๐๐ is very similar to . OC Filter Opening followed by Closing In [75-76] a switching VMF based on quaternion impulse ๐๐ ๐๐ CO Filter Closing followed by Opening (79) ๐๐ ๐๐ ร 5 detector is proposed using 5 window. The quaternion Erosion distributesโ the local minima while dilation color difference between pixels along the four directional โ ยฐ ยฐ ยฐ ยฐ distributes the local maxima within the sliding window operators at0 , 45 , 90 and135 are computed. An average defined by the structuring element which resembles a color difference between the central pixel and other min/max filtering action for suppression of impulse noise. pixel values in the direction = (1, 2, 3,4) for respective In [86], an extension of MF to multichannel images on ๐๐๐๐๐๐โ degrees are computed as follows: learning-based morphological color operations is = , โ (77) developed. The color pixel ordering scheme is learned in = min1 (4 , , , ) accordance to pre-estimation of healthy and corrupted โ 4 ๐๐=1 ๐๐ ๐๐ The๐๐๐๐๐๐ aboveโ two โฃequations๐๐๏ฟฝ๐๐ ๐๐ ๏ฟฝ areโฃ used for impulse detection. If pixels where the corrupted samples are ordered either as 1 2 3 4 ๐๐the๐๐๐๐ central pixel๐๐๐๐๐๐ is๐๐ ๐๐an๐๐ impulse,๐๐๐๐๐๐ ๐๐๐๐ its๐๐ value will be large maximum in erosion or minimum in dilation. The SVM is otherwise it is small. Hence the output of the filter is given used as the classification technique of such learning-based by ๐๐๐๐๐๐ multichannel MF in the RGB color domain. After each ( ), if morphological operation, the image reconstruction step is = (78) carried out for restoring the original features. ( ๐๐๐๐๐๐)/ , otherwise ๐๐๐๐๐๐๐๐ ( ๐๐) ๐๐๐๐๐๐ โฅ ๐๐ Extension of mathematical morphology to multivariable w๐๐ here ๏ฟฝ represents the VMF calculated in Quaternion ๐๐ ๐๐+1 2 data such as multichannel image is performed in [87-89]. form and ๐๐๐๐๐๐T being a pre-specified threshold. In order to develop multivariate morphological operators, ๐๐ In [77] another Quaternion Switching Vector Median various complex mathematical tools such as machine Filter (QSVMF) is developed based on both the intensity learning [90], principal component analysis [91], hyper and chromaticity differences described above. complex [92], rand projection depth function [93], group- A modification of QSVMF is designed in [78]. It is invariant frames [94], and probabilistic extrema estimation different from other related works in the impulse detection [95] are used. which utilizes the pixels along only a direction with Based on the self-duality property of morphological maximum number of similar pixels while other utilize the operator, a multivariate self-dual morphological operator color differences between the central pixel and its is developed in [96] which is applicable for noise removal neighboring pixels in the four-edge direction. and segmentation purposes. A pair of symmetric vector A two-stage filter using both the Quaternion based Switching filter and a local mean filter is designed in [79] 78 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 ordering is introduced in order to develop multivariate remaining pixels which are not flagged as outliers are used dual morphological operators. for calculating the Vector median and the noisy pixel is Quantile filters and a group of morphological gradient replaced by the corresponding vector median. filters are proposed in [97]. Their properties and links with dilation and erosion operators are also investigated. 2.12.2 Similarity based Impulsive Noise Removal Filter Another quantile filter is also designed in [98]. In [99], a new hybrid filter based on mathematical In [106] a filter based on similarities between the pixels in morphology and trimmed standard median filter is a predefined window is developed. If { , โฆ , } denote the pixels in a filtering window, then a convex similarity developed. For impulse detection, mathematical ๐๐ ๐๐ morphology such as erosion and dilation are utilized. function is used to find the similarity between๐๐ ๐๐pixels. For Erosion refers to the computation of local minima while the central pixel and its neighboring pixels, the cumulated dilation estimates the local maxima of the neighboring sum of similarities are computed as follows: pixels. The existence of an inverse relationship between = ( , ), ๐๐ ๐๐ erosion and dilation is used for detection for salt & pepper = , ( , ) (82) 1 โ๐๐=2 1 ๐๐ noise. ๐๐ ๐๐ ยต ๐๐ ๐๐ ( , ) ๐๐is the ๐๐=central2 ๐๐โ ๐๐ pixel๐๐ ๐๐and is a convex similarity ๐๐ โ ยต ๐๐ ๐๐ < function.1 The central pixel is ๐๐detected๐๐ as noisy if 2.11 Wavelet-Transform based Filters ๐๐ , = 2, โฆ , and is replacedยต ๐๐ ๐๐by that for which = 1 , = 2, โฆ , .This filter is faster than VMF.๐๐ A Many nonlinear filtering techniques are designed based on ๐๐ ๐๐ similar๐๐ ๐๐ filter based๐๐ on this idea is also developed๐๐ in [107๐๐]. wavelet transform. Haci Tasmaz et al., [100] proposes a ๐๐๐๐๐๐๐๐๐๐๐๐๐๐๐๐ ๐๐ ๐๐ satellite image enhancement system comprising of image 2.12.3 Filters based on Digital Path Approach denoising and illumination enhancement technique. It is based on dual tree complex wavelet transform (DT-CWT). In [108] noise filtering approach based on the connections Based on the combined effect of Haar wavelet transform between image samples using the digital geodesic path is and median filter, a denoising technique is also proposed developed instead of using a fixed window. In this filter [101]. In [102], another denoising algorithm based on the image pixels are grouped togetherthrough the fuzzy combined effect of the bi-orthogonal wavelet transform membership function defined over vectorial inputs by and median filter is designed which removes noise forming digital paths revealing the underlying structure effectively. In [103], a wavelet based denoising technique dynamics of the image. It shows better performance in for suppression of noise in Magnetic Resonance images suppressing impulsive, Gaussian as well as mixed-type (MRI) is proposed. Shalini and Godwin in [104] present a noise. comparative analysis of denoising algorithms based on different wavelet transform such as Bior, Surelet, Haar 2.12.4 Filters based on Long Range Correlation and Curvelet. Traditional method of noise filtering utilizes information 2.12 Miscellaneous Filters based on local window centered around the corrupted pixel. In [109] a new class of filter based on information Miscellaneous Filters consist of those filters which cannot on both the local window and also some remote regions in be included into any of the filter families described above the image is developed. This is due to the fact that there although some filters may have common properties. exists a long range correlation within natural images and also the human visual system can use such correlations to 2.12.1 Selective Vector Median Filter interpret and restore image information [110]. This filter comprises of two steps: impulse detection and noise In [105] Selective Vector Median Filter is introduced cancellation. In the impulse detection stage any impulse which has two steps: noise detection and noise removal. detector as described in [111] and [112] can be used for For every pixel in the window aggregated sum of distance identifying the corrupted pixel in the local window. If the between other pixels is computed central pixel is corrupted, then a remote window centered = , = 1, โฆ (80) around the impulse pixel is defined in the search range Then the๐๐ mean value for that neighborhood is calculated which is larger than the local window. All the pixels in the ๐๐๐๐ โ๐๐=1 โฅ ๐๐๐๐ โ ๐๐๐๐ โฅ ๐๐ ๐๐ as follows remote window will compete for the perfect match with = ๐๐ (81) the local window based on the mean square error of The 1pixels๐๐ which have value higher than in the uncorrupted pixels in local and remote window. Finally, ฬ ๐๐ โ๐๐=1 ๐๐ ๐๐neighborhood๐๐ are flagged as outliers. represents a the corrupted pixel is replaced by the central pixel of the constant used to increase or๐๐ decrease the sensitivity๐๐๐๐ฬ of the remote window with minimum mean squared error. threshold. If the central pixel is an outlier,๐๐ then the IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 79
Fig. 5 Calculation of similarity measure between pixel in the processing block (Green Square) and small window (red square) by ROAD. Fig. 4 Demonstration of the general approach 2.12.7 Vector Marginal Median Filters (VMMF) 2.12.5 Vector Rational Filters (VRF) In [3] a vector filter named Vector Marginal Median Filter Vector rational filters are extension of rational filters. In (VMMF) based on the median operation is presented. It [113], two algorithms are proposed and based calculates the median value of each channel separately and on -norm and directional processing using two the central pixel is replaced by the median value of the ๐ฟ๐ฟ ๐๐ decimation schemes such as rectangular๐๐๐๐๐๐ and๐๐๐๐๐๐ quincunx respective channel. The output of VMMF is given below ๐๐ decimation๐ฟ๐ฟ for down-sampling. Merits of these filters are = (( ({ , โฆ })), ({ , โฆ }) , ( ({ , โฆ }))) better edge-preserving properties and absence of artifacts. ๐๐๐๐๐๐๐๐ ๐๐ ๐ ๐ ๐ ๐ ๐บ๐บ ๐บ๐บ ๐ต๐ต ๐ต๐ต (85) 1 ๐๐ 1 ๐๐ 1 ๐๐ 2.12.6 Robust Local Similarity Filters (RLSF) where๐๐๐๐๐๐ ๐ฅ๐ฅ, and๐ฅ๐ฅ are๏ฟฝ๐๐ ๐๐๐๐the red,๐ฅ๐ฅ green๐ฅ๐ฅ ๏ฟฝ and๐๐๐๐๐๐ blue๐ฅ๐ฅ channel๐ฅ๐ฅ of the pixels in the window. Its noise reduction capability is A new algorithm for suppression of mixed Gaussian and highest๐ ๐ but๐บ๐บ since๐ต๐ต it does not consider the correlation impulsive noise based on concept of Rank ordered among the vector channel, it leads to color artifacts. Absolute Difference (ROAD) [114] is proposed in [115]. Noise is detected by computing ROAD between samples 2.12.8 Vector Signal-Dependent Rank Order Mean Filters of the processing block and a small window which is centered at blockโs central pixel. Main contribution is that Multichannel extension of the grayscale Signal Dependent the similarity measure is not affected by outliers due to Rank Order Mean (SDROM) filter [116] is the Vector impulses and the averaging operation reduces the signal-dependent rank order mean (VSDROM) filter [117]. Gaussian noise. The output of this filter is given as For noise detection, first the vector pixels in the window . are sorted according to their aggregated distance to all = ๐๐ (83) other samples. The distance between the four lowest- โ๐๐=1 ๐ค๐ค๐๐ ๐๐๐๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐ ๐๐ ranked vectors and the central pixel are compared against ๐๐with =โ๐๐=1 ๐ค๐ค๐๐ ( ) (84) respective increasing thresholds. The central pixel is 1 ๐ผ๐ผ ๐๐ ๐๐=1 ๐๐ regarded as noisy if any of these distances is greater than ๐ค๐ค า ๏ฟฝ๐ผ๐ผ โ ๐๐ ๐๐ ๏ฟฝ where represents the kernel function (Gaussian) and their threshold and replaced by the VMF. ( ) denotes the smallest Euclidean distance า ๐ก๐กโ 2.12.9 Decision based Couple Window Median Filters between๐๐ pixel of the processing block and pixels of ๐๐ ๐๐ ๐๐ (DBCWMF) of the small window.๐๐ is the number of close neighbors used in calculation๐ฅ๐ฅ of ROAD. ๐๐ ๐ผ๐ผ They use an increasing window size in order to maximize the probability of finding noise-free pixels [118]. The steps of DBCWMF is given below 1. Define a 2-D local window of dimension (2 + 1) ร (2 + 1) (initialize algorithm by choosing = 1). ๐๐ 2. Identify salt & pepper noise by๐๐ checking if the central๐๐ pixel ๐๐ is either 0 or 255. If 0< < 255๐๐ it is left intact. ๐๐+1 ๐๐+1 ๐๐ 2 ๐๐ 2 80 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016
3. If is noisy, then trim all 0โs and 255โs present in uncontaminated, ( = 1,2,3) equals 0 or 255 with equal probability for fixed-valued impulse noise, or takes any the ๐๐and+1 follow the two cases ๐๐ ๐๐ 2 value in the range๐๐ ๐๐[0,255] for random-valued impulse Case 1. If the number of non-noisy pixels is non-zero, then ๐๐ noise; is the sample corruption probability; , and replace๐๐ the central pixel by the median value calculated are the channel corruption probabilities and = 1 from the remaining uncorrupted pixels. 1 2 3 ๐๐ . ๐๐ ๐๐ ๐๐ Case 2. If the number of non-noisy pixels is zero, then ๐๐๐๐ โ = + 1, ( < 5) The uncorrelated impulse noise has the following form [46] update the window size by increasing ๐๐1 โ ๐๐2 โ ๐๐ 3 with probability and goto Step 1. = with probability 1 (87) 4. If the number of non-noisy pixels ๐๐in ๐๐ is zero,๐๐ then โฒ ๐๐ ๐๐ ๐๐ = 1,2,3 ๐๐ w๐๐ here๏ฟฝ denotes the three channels in RGB color replace the value of central pixel by mean of . ๐๐๐๐ โ ๐๐ ๐๐4 space; is the channel corruption probability; and 5. Repeat Steps 1 to 4 until all the pixels are processed. ๐๐ ๐๐1 denote the original component and contaminated ๐๐ ๐๐๐๐ ๐๐๐๐ 2.12.10 Improved Bilateral Filter for reducing mixed component respectively. can take either 0 or 255 for fixed-valued impulse noise and can take any discrete value noise ๐๐ in [0,255] for random-valued๐๐ impulse noise. A Bilateral Filter (BF) [119] is a combination of two low- pass Gaussian filters operating in spatial and color (range) domain for reducing both impulse and additive noise 4. Performance Measurement of Filters while preserving edge structures. The spatial low-pass The performance of filter is evaluated by the following filter gives larger weights to those samples closer to the parameters central pixel while the range low pass filter assigns larger (a) Mean Absolute Error (MAE) weights to those pixels having similar color distributions with the central pixel. Here the filterโs output is mainly = | ( , ) ( , )| (88) dependent on the central pixel and neighboring pixels ร 1 ๐๐1 ๐๐1 close in both spatial and range domain with the central where ร is the size of the image; ( , ) and ( , ) ๐๐๐๐๐๐ ๐๐1 ๐๐1 โ๐๐=1 โ๐๐=1 ๐๐ ๐๐ ๐๐ โ ๐๐ ๐๐ ๐๐ pixel.An improvement in the traditional Bilateral Filter are the original and filtered pixel values at ( , ) location. 1 1 (BF) is designed in [120] in which a new weighting (b) Mean๐๐ Squared๐๐ Error (MSE) ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ ๐๐ function to the bilateral filtering mechanism is introduced. = ( ( , ) ( , )) ๐๐ ๐๐ (89) ร For an impulse, the vector median (as opposed to the 1 ๐๐1 ๐๐1 (C) Peak Signal to Noise ratio (PSNR) 2 traditional BF which always uses the central pixel) is ๐๐๐๐๐๐ ๐๐1 ๐๐1 โ๐๐=1 โ๐๐=1 ๐๐ ๐๐ ๐๐ โ ๐๐ ๐๐ ๐๐ considered as base for bilateral filtering operation for = 10log (90) 2 ๐๐๐๐๐๐ replacement of the central pixel otherwise the normal where is the ๐ผ๐ผmaximum pixel value of the original 10 ๐๐๐๐๐๐ bilateral filtering action is continued. image.๐๐๐๐๐๐๐๐ ๏ฟฝ ๏ฟฝ ๐๐๐๐๐๐ (d) Normalized๐ผ๐ผ Color Distance (NCD) The NCD is defined in the Lu*v* color space by 3. Impulse Noise Model =
( , ) ( , ) ( , ) ( , ) ( , ) ( , ) In the real life, the form of impulse noise varies. For ๐๐๐๐๐๐ ๐๐1 ๐๐1 ๐๐ ๐ฅ๐ฅ 2 ๐๐ ๐ฅ๐ฅ 2 ๐๐ ๐ฅ๐ฅ 2 example, the value of impulse noise caused by โ๐๐=1 โ๐๐=1 ๏ฟฝ๏ฟฝ๐ฟ๐ฟ ๐๐ ๐๐ โ๐ฟ๐ฟ ๐๐ ๐๐ ๏ฟฝ( ,+)๏ฟฝ๐ข๐ข ๐๐ ๐๐ โ( ๐ข๐ข, ) ๐๐ ๐๐ ๏ฟฝ +(๏ฟฝ๐ฃ๐ฃ, ) ๐๐ ๐๐ โ๐ฃ๐ฃ ๐๐ ๐๐ ๏ฟฝ malfunction of sensor is expected to be fixed-valued ๐๐ ๐๐ 2 2 2 (91) 1 1 ๏ฟฝ ๐๐ ๐๐ ๐๐ impulse noise, while the value of impulse caused by โ๐๐=1 โ๐๐=1 ๏ฟฝ๐ฟ๐ฟ ๐๐ ๐๐ ๏ฟฝ +๏ฟฝ๐ข๐ข ๐๐ ๐๐ ๏ฟฝ +๏ฟฝ๐ฃ๐ฃ ๐๐ ๐๐ ๏ฟฝ ( , ), ( , ) ( , ) ( , ), ( , ) electronic interference is random-valued impulse noise where , and , ( ) ๐๐ ๐๐ ๐๐ ๐ฅ๐ฅ ๐ฅ๐ฅ [77]. Impulse noise can be divided into two types: , are values of the lightness and two chrominance ๐ฟ๐ฟ ๐๐ ๐๐ ๐ข๐ข ๐๐ ๐๐ ๐ฃ๐ฃ ๐๐ ๐๐ ๐ฟ๐ฟ ๐๐ ๐๐ ๐ข๐ข ๐๐ ๐๐ Correlated impulse noise and Uncorrelated impulse noise. components๐ฅ๐ฅ of the original image sample ( , ) and ๐ฃ๐ฃ ๐๐ ๐๐ The impulse noise model proposed by Viero et al [10] is filtered image sample ( , ) respectively. ๐๐ ๐๐ ๐๐ correlated type impulse noise and it has the following (E) Structural Similarity Index (SSIM) ๐๐ ๐๐ ๐๐ form Structural similarity index measures the similarity { }with probability 1 between two images and is given below , , { 1 2 3 }with probability = (92) ๐๐ ๐๐ ๐๐ ๐ฅ๐ฅ ๐ฆ๐ฆ 1 ๐ฅ๐ฅ๐ฅ๐ฅ 2 = { , , }with probability โ ๐๐ ๏ฟฝ2๐๐ ๐๐ +๐ถ๐ถ ๏ฟฝ๏ฟฝ2๐๐ +๐ถ๐ถ ๏ฟฝ โง 1 2 3 1 (86) 2 2 2 2 ๐๐ ๐๐ ๐๐ where ยตx๏ฟฝ ๐๐and๐ฅ๐ฅ+๐๐๐ฆ๐ฆ +ยต๐ถ๐ถy1 ๏ฟฝare๏ฟฝ๐๐๐ฅ๐ฅ+ ๐๐mean๐ฆ๐ฆ+๐ถ๐ถ2๏ฟฝ of the original and filtered โฒ โช { , , }with probability ๐๐ ๐๐ ๐๐๐๐๐๐๐๐ 1 2 3 2 2 2 image, ฯxy, ฯ and ฯ represent the corresponding ๐๐ {๐๐ , ๐๐ , ๐๐ }with probability ๐๐ ๐๐ x y โจ 1 2 3 3 where ๐๐= ( ๐๐ , ,๐๐ ) is the original๐๐ uncontaminated๐๐ covariance and variance of the original and filtered images. โช 1 2 3 ๐๐ vectorโฉ ๐๐pixel,๐๐ ๐๐ may be either ๐๐contaminated๐๐ or and are the constants. 1 2 3 ๐๐ ๐๐ ๐๐โฒ ๐๐ 1 2 ๐๐ ๐ถ๐ถ ๐ถ๐ถ IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 81
MAE is used to evaluate detail preservation; MSE and [14] K. M. Singh, and P.K. Bora, โAdaptive vector median filter PSNR are used to evaluate noise suppression capability; for removal of impulse noise from color imagesโ, Journal of NCD is used to measure perceptual error in the CIELab Scientific Computing, vol. 4, no.1, pp.1063-1072, 2004. color space [3]. [15] B. Smolka, K.N. Plataniotis, and A.N. Venetsanopoulos, โNonlinear techniques for color image processingโ, For an efficient filter, it is expected to have high PSNR Nonlinear Signal and Image Processing: Theory, Methods, and SSIM, while the other parameters like MAE, MSE and Applications, pp.445-505, 2004. and NCD are minimum. [16] K. M. Singh, and P.K. Bora, โSwitching vector median filters based on non-causal linear prediction for detection of impulse noiseโ, The Imaging Science Journal, vol.62, no.6, 5. Conclusions pp.313-326, 2014. [17] P. E. Trahanias, and A. N. Venetsanopoulos, โVector A comprehensive survey of various vector median filters directional filters-a new class of multichannel image for the removal of impulse noise from color images is processing filtersโ, IEEE Transactions on Image Processing, presented. These filters have been categorized into 12 vol.2, no.4, pp.528-534, 1993. different families. Their properties have been summarized [18] P. E. Trahanias, D. Karakos, and A. N. Venetsanopoulos, โDirectional processing of color images: theory and and presented. Some recently proposed algorithms have experimental resultsโ, IEEE Transactions on Image been added and studied. Processing, vol.5, no.6, pp. 868-880, 1996. [19] D. G. Karakos, and P. E. Trahanias, โCombining vector References median and vector directional filters: The directional- [1] H. J. Trussell, E. Saber, and M. Vrhel, โColor image distance filtersโ, Image Processing, Proceedings. processing: Basics and special issue overviewโ, IEEE International Conference on. Vol. 1. IEEE, 1995. Signal Processing Mag., vo. 22, no.1, pp. 14-22, 2005. [20] G. P. Deepti, M. V. Borker, and J. Sivaswamy, โImpulse [2] T. Acharya, and K. R. Ajay, Image processing: principles noise removal from color images with Hopfield neural and applications. John Wiley & Sons, 2005. network and improved vector median filterโ, Sixth Indian [3] K. Plataniotis, and A. N. Venetsanopoulos, Color image Conference on Computer Vision, Graphics & Image processing and applications. Springer, 2000. Processing, 2008. [4] I. Pitas, and A. N. Venetsanopoulos, Nonlinear Digital [21] D. G. Karakos, and P. E. Trahanias, โGeneralized Filters Principles and Applications. Kluwer Academic multichannel image-filtering structuresโ, IEEE Transactions Publishers, Boston, M.A, 1990. on Image Processing, vol. 6, no.7, pp. 1038-1045, 1997. [5] R. Lukac, and K. N. Plataniotis, Color image processing: [22] R. Wichman, K. Oistamo, Q. Liu, M. Grundstrom, and Y. A. methods and applications. CRC press, 2006. Neuvo, โWeighted vector median operation for filtering [6] M. E. Celebi, H. A. Kingravi, and Y. A. Aslandogan, multispectral dataโ , Applications in Optical Science and โNonlinear vector filtering for impulsive noise removal Engineering. International Society for Optics and Photonics, from color imagesโ, Journal of Electronic Imaging, vol.16, 1992. no.3, 2007. [23] B. Smolka, โRank Order Weighted Vector Median Filterโ, [7] R. Rastislav, and K. N. Plataniotis, โA taxonomy of color 17th International Conference on Systems, Signals and image filtering and enhancement solutionsโ, Advances in Image Processing, IWSSIP 2010. imaging and electron physics, vol.140, 2006. [24] B. Smolka, โRank-Based Vector Median Filterโ, [8] R. Lukac, and S. Marchevsky, โImpulse detection in noised Proceedings: ASIPA ASC 2009: Asia Pacific Signal and colour imagesโ, Journal of Electrical Engineering, vol. 52, Information Processing Association, 2009 Annual Summit pp. 352-361, 2001. and Conference: 254-257. [9] J. Astola, P. Haavisto, and Y. Neuvo, โVector median [25] R. Lukac, K. N. Plataniotis, B. Smolka, and A. N. filtersโ, Proceedings of the IEEE, vol. 78, no.4, pp. 678-689, Venetsanopoulos, โSelection weighted vector directional 1990. filtersโ, Computer Vision and Image Understanding, vol.94, [10] T. Viero, K. Oistamo, and Y. Neuvo, โThree-dimensional no.1, pp.140-167, 2004. median-related filters for color image sequence filteringโ, [26] R. Lukac, K. N. Plataniotis, B. Smolka, and A. N. IEEE Transactions on Circuits and Systems for Video Venetsanopoulos, โGeneralized selection weighted vector Technology, vol.4, no. 2, pp.129-142, 1994. filtersโ, EURASIP Journal on Advances in Signal [11] B. Smolka, and M. Perczak, โGeneralized vector median Processing , vol.12, pp.1-16, 2004. filterโ, Image and Signal Processing and Analysis, 2007. [27] R. Lukac, K. N. Plataniotis, and A.N. Venetsanopoulos, ISPA 2007. โColor image denoising using evolutionary computationโ, [12] B. Smolka, M. Szczepanski, K. N. Plataniotis, and A. N. International Journal of Imaging Systems and Technology, Venetsanopoulos, โFast modified vector median filterโ, in vol.15, no.5, pp. 236-251 , 2005. W.Skarbek (Ed.) Computer Analysis of Images and Patterns, [28] S. G. Javed, A. Majid, A. M. Mirza, and A. Khan, โMulti- LNCS 2124, Springer-Verleg-Berlin, pp. 570-580, 2001. denoising based impulse noise removal from images using [13] S. Vinayagamoorthy, K. N. Plataniotis, D. Androutsos, and robust statistical features and genetic programmingโ, vol.75, A. N. Venetsanopoulos , โA multichannel filter for TV no.10, pp.1-30, 2015. signal processingโ, IEEE transactions on consumer [29] R. Lukac, โAdaptive color image filtering based on center- electronics, vol.42, no.2, pp. 199-202, 1996. weighted vector directional filtersโ, Multidimensional 82 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016
Systems and Signal Processing, vol.15, no.2, pp.169-196, [47] B. Smolka, and A. Chydzinski, โFast detection and 2004. impulsive noise removal in color imagesโ, Real-Time [30] Z. Ma, H. R. Wu, and D. Feng, โPartition-based vector Imaging, vol.11, no.5, pp.389-402, 2005. filtering technique for suppression of noise in digital color [48] L. Malinski, and B. Smolka, โFast averaging peer group imagesโ, IEEE transactions on image processing, vol.15, filter for the impulsive noise removal in color imagesโ, no.8, pp.2324-2342, 2006. Journal of Real-Time Image Processing, vol.11, no.3, [31] R. Lukac, โAdaptive vector median filteringโ, Pattern pp.427-444, 2016. Recognition Letters, vol. 24, no.12, pp.1889-1899, 2003. [49] S. Morillas, V. Gregori, and A. Hervรกs, โFuzzy peer groups [32] B. Smolka, and K. Wojciechowski, โOn the adaptive for reducing mixed Gaussian-impulse noise from color impulsive noise removal in color imagesโ, Archives of imagesโ, IEEE Transactions on Image Processing, vol.18, Control Sciences, vol.15, no.1, pp.117-131, 2005. no.7, pp.1452-1466, 2009. [33] A. Roy, and R. H. Laskar, โMulticlass SVM based adaptive [50] J. G. Camarena, V. Gregori, S. Morrilas, and A. Sapena, filter for removal of high density impulse noise from color โFast detection and removal of impulsive noise using peer imagesโ, Applied Soft Computing, vol.46, pp. 816-826, groups and fuzzy metricsโ, Journal of Visual 2015. Communication and Image Representation, vol.9, no.1, [34] U. Ghanekar, A. K. Singh, and R. Pandey, โRandom Valued pp.20-29, 2008. Impulse Noise Removal in Colour Images using Adaptive [51] R. Lukac and K. N. Plataniotis, โFuzzy Vector Filters for Threshold and Colour Correctionโ, ACEEE Int. J. on Signal cDNA Microarray Image Processingโ, Studies in Fuzziness & Image Processing, vol.1, no.3, 2010. and soft computing, vol. 242, pp.67-82, 2008. [35] M. E. Celebi, G. Schaefer, and H. Zhou, โA new family of [52] K. N. Plataniotis, D. Androutsos, and A. N. order-statistics based switching vector filtersโ, 17th IEEE Venetsanopoulos, โAdaptive fuzzy systems for International Conference on Image Processing on Image multichannel signal processingโ, Proc. IEEE, vol. 87, no.9, Processing, 2010. pp. 1601-1622, 1999. [36] M. E. Celebi, and Y. A. Aslandogan, โRobust switching [53] K. N. Plataniotis, D. Androutsos, and A. N. vector median filter for impulsive noise removalโ, Journal Venetsanopoulos, โFuzzy adaptive filters for multichannel of Electronic Imaging, vol.17, no.4, 2008. image processingโ, Signal Processing Journal, vol.55, [37] S. Morillas, V. Gregori, and A. Sapena., โAdaptive no.1,pp. 93-106, 1996. marginal median filter for colour imagesโ, vol. 11, no. 3, [54] P. T. Yu, and R. C. Chen, โFuzzy Stack Filters - Their pp.3205-3213, Sensors, 2011. definitions, fundamental properties and application in image [38] R. Lukac, โOptimised directional distance filterโ, Machine processingโ, IEEE Transactions on Image Processing, vol.5, Graphics & Vision International Journal, vol.11, no.2/3, pp. no.6, pp.838-54, 1996. 311-326, 2002. [55] S. Peng, and L. Lucke, โFuzzy filtering for mixed noise [39] R. Lukac, and B. Smolka, โApplication of the adaptive removal during image procesisngโ, Fuzzy Systems, 1994. center-weighted vector median framework for the IEEE World Congress on Computational Intelligence. enhancement of cDNA microarray imagesโ, International Proceedings of the Third IEEE Conference on. IEEE, 1994. Journal of Applied Mathematics and Computer Science, [56] K. N. Plataniotis, D. Androutsos, S. Vinayagamoorthy, and vol.13, no.3, pp. 369-384, 2003. A. N. Venetsanopoulos, โA nearest neighbor multichannel [40] S. Wang, and C. H. Wu, โA new impulse detection and filterโ, Electron. Lett. vol.31, no.22, pp.1910-1911, 1995. filtering method for removal of wide range impulse noiseโ, [57] K. N. Plataniotis, D. Androutsos, S. Vinayagamoorthy, and Pattern Recognition, vol.42, no.9, pp. 2194-2202, 2009. A. N. Venetsanopoulos, โAn adaptive nearest neighbor [41] R. Lukac, B. Smolka, and K. N. Plataniotis, โSharpening multichannel filterโ, IEEE Trans. Circuits Syst. Video vector median filtersโ, Signal Processing, vol. 87, no.9, Technology., vol. 6, no.6, pp. 699-703 1996. pp.2085-2099, 2007.. [58] H. T. Tsai and P. T. Yu, โAdaptive fuzzy hybrid [42] B. Smolka, K. Malik, and D. Malik, โAdaptive rank multichannel filters for removal of impulse noise from color weighted switching filter for impulsive noise removal in imagesโ, Signal Processing, vol.74, pp.127-151, 1999. color imagesโ, Journal of Real-Time Image Processing, [59] S. Schulte, M. Nachtegael, V. De Witte, D. Van der Weken, vol.10, no.2, pp.289-311, 2015. and E. E. Kerre , โA fuzzy impulse noise detection and [43] Y. Deng, C. Kenney, M. S. Moore, and B. S. Manjunath, reduction methodโ, IEEE Transactions on Image Processing, โPeer group filtering and perceptual color image vol. 15, no.5, pp.1153-1162, 2006. quantizationโ, Circuits and Systems, ISCAS'99. [60] L. Khriji, and M. Gabbouj, โVector median-rational hybrid Proceedings of the 1999 IEEE International Symposium on. filters for multichannel image processingโ, IEEE Signal Vol. 4. IEEE, 1999. Processing Letters, vol. 6, no.7, pp.186-190, 1999. [44] C. Kenney, Y. Deng, B. S. Manjunath, and G. Hewer, โPeer [61] Z. Ma, H. R. Wu, and B. Qiu, โA robust structure-adaptive group image enhancementโ, IEEE Transactions on Image hybrid vector filters for color image restorationโ, IEEE Processing, vol. 10, no.2, pp. 326-334 , 2001. Transactions on Image Processing, vol.14, no.12, pp.1990- [45] J. Y. Ho, โPeer region determination based impulsive noise 2001, 2005. detectionโ, Acoustics, Speech, and Signal Processing, 2003. [62] L. Khriji, and M. Gabbouj, โClass of multichannel image [46] B. Smolka, โPeer group switching filter for impulse noise processing filtersโ, Electronics Letters, vol.35, no.4, pp. reduction incolor imagesโ, Pattern Recognition Letters, 285-287, 1999. vol.31, no.6, pp. 484-495, 2010. [63] R. Lukac, B. Smolka, K. N. Plataniotis, A. N. Venetsanopoulos, and P. Zavarsky, โAngular multichannel IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016 83
sigma filterโ, Acoustics, Speech, and Signal Processing, sequencesโ, IEEE Transactions on Circuits and Systems for vol.3, 2003. Video Technology, vol. 23, no.5, pp.741-755, 2013. [64] R. Lukac, B. Smolka, K. N. Plataniotis, and A. N. [82] G. Elaiyaraja, N. Kumaratharan, and C. R. Prapau, Venetsanopoulos, โVector sigma filters for noise detection โModified Quaternion Based Impulse Noise Removal with and removal in color imagesโ, Journal of Visual Adaptive Threshold from Color Video Sequences and Communication & Image Representation, vol. 17, pp. 1-26, Medical Imagesโ, Middle-East Journal of Scientific 2006. Research, vol. 23, no.7, pp. 1382-1389, 2015. [65] R. Lukac, B. Smolka, K. N. Plataniotis, and A. N. [83] K. Michielsen, and H. D. Raedt, โMorphological image Venetsanopoulos, โA variety of multichannel sigma analysisโ, Computer Physics Communications, vol. 132, filtersโ, Optical Metrology, International Society for Optics no.1, pp. 94-103 , 2000. and Photonics, 2003. [84] J. Serra, Image analysis and mathematical morphology, [66] J.S Lee, โDigital image smoothing and the sigma Academic press, 1982. filterโ, Computer Vision, Graphics, and Image Processing, [85] R. Stevenson, and G. Arce, โMorphological filters: vol. 24, no.2, pp.255-269, 1983. Statistics and further syntactic propertiesโ, IEEE [67] R. Lukac, B. Smolka, K. N. Plataniotis, and A. N. Transactions on Circuits and Systems, vol.34, no.11, pp. Venetsanopoulos, โGeneralized adaptive vector sigma 1292-1305, 1987. filtersโ, Multimedia and Expo, 2003. [86] H. Zhou, and K. Z. Mao, โAn impulsive noise color image [68] A. Beghdadi, and A. Khellaf, โA noise-filtering method filter using learning-based color morphological using a local information measureโ, IEEE Transactions on operationsโ, Digital Signal Processing, vol.18, no.3, pp. Image Processing, vol.6, no.6, pp.879-882, 1997. 406-421, 2008. [69] R. Lukac, B. Smolka, K. N. Plataniotis, and A. N. [87] S. V. Forero and J. Angulo, โClassification of hyperspectral Venetsanopoulos, โEntropy vector median filterโ, Iberian images by tensor modeling and additive morphological Conference on Pattern Recognition and Image Analysis, decompositionโ, Pattern Recognition, vol. 46, no. 2, pp. Springer Berlin Heidelberg, 2003. 566โ577, 2013. [70] R. Lukac, K. N. Plataniotis, B. Smolka, and A. N. [88] M. E. Valle and D. M. G. Vicente, โSparsely connected Venetsanopoulos, โGeneralized entropy vector filtersโ, autoassociative lattice memories with an application for the Video/Image Processing and Multimedia Communications, reconstruction of color images,โ Journal of Mathematical 2003. Imaging and Vision, vol. 44, no. 3, pp. 195โ222, 2012. [71] Y.B Jia, โQuaternions and rotationsโ, Com S 477.577, 2008. [89] S. Morales, V. Naranjo, U. Angulo, and M. Alcaniz, [72] J. Angulo, โQuaternion colour representations and derived โAutomatic detection of optic disc based on PCA and total orderings for morphological operatorsโ, Conference on mathematical morphology,โ IEEE Transactions on Medical Colour in Graphics, Imaging, and Vision, vol. 2008, no.1, Imaging, vol. 32, no. 4, pp. 786โ796, 2013. pp.417-422, 2008. [90] S. Velasco-Forero and J. Angulo, โSupervised ordering [73] T.A Ell, and S.J. Sangwine, โTheory of vector filters based in IRP: application to morphological processing of on linear quaternion functionsโ, Signal Processing hyperspectral images,โ IEEE Transactions on Image Conference, IEEE, 2008. Processing, vol. 20, no. 11, pp. 3301โ3308, 2011. [74] C. Cai, and S.K. Mitra, โA normalized color difference edge [91] J. Li and Y. Li, โMultivariate mathematical morphology detector based on quaternion representationโ, Image based on principal component analysis: initial results in Processing, 2000. building extraction,โ in Proceedings of the 20th [75] L.Jin, and D. Li, โAn efficient color-impulse detector and International Society for Photogrammetry and Remote its application to color imagesโ, IEEE Signal Processing Sensing (ISPRS '04), vol. 35, pp. 1168โ1173, 2004. Letters, vol.14, no.6, pp.397-400, 2007. [92] J. Angulo, โHypercomplex mathematical [76] X. Zhang, โA Novel Switching Vector Median Filter for morphology,โ Journal of Mathematical Imaging and Vision, Color Image Filteringโ, Informatics and Management vol. 41, no. 1-2, pp. 86โ108, 2011. Science V, pp. 663-670, Springer London, 2013. [93] S. Velasco-Forero and J. Angulo, โRandom projection [77] X. Geng, X. Hu, and J. Xiao, โQuaternion switching filter depth for multivariate mathematical morphology,โ IEEE for impulse noise reduction in color imageโ, Signal Journal on Selected Topics in Signal Processing, vol. 6, no. Processing, vol. 92, no.1, pp. 150-162, 2012. 7, pp. 753โ763, 2012. [78] Jin, Lianghai, H. Liu, X. Xu, and E. Song, โColor impulsive [94] J. J. van de Gronde and J. B. Roerdink, โGroup-invariant noise removal based on quaternion representation and frames for colour morphology,โ in Proceedings of the directional vector order-statisticsโ, Signal Processing, International Symposium on Memory Management (ISMM vol.91, no.5, pp. 1249-1261, 2011. '13), pp. 267โ278, May 2013. [79] G. Wang, Y. Liu, and T. Zhao, โA quaternion-based [95] A. Cฤliman, M. Ivanovici, and N. Richard, โProbabilistic switching filter for colour image denoisingโ, Signal pseudo-morphology for grayscale and color Processing , vol.102, pp. 216-225, 2014. images,โ Pattern Recognition, vol. 47, no. 2, pp. 721โ735, [80] Jin, Lianghai, Z. Zhu, X. Xu, and X. Li, โTwo-stage 2014. quaternion switching vector filter for color impulse noise [96] T. Lei, Y. Wang, and W. Luo, โMultivariate Self-Dual removalโ, Signal Processing, vol. 128, pp. 171-185, 2016. Morphological Operators Based on Extremum [81] Jin, Lianghai, H. Liu, X. Xu, and E. Song, โQuaternion- Constraintโ, Mathematical Problems in Engineering, 2015. based impulse noise removal from color video [97] M. Welk, A. Kleefeld, and M. Breuร, โQuantile filtering of colour images via symmetric matricesโ, Mathematical 84 IJCSNS International Journal of Computer Science and Network Security, VOL.16 No.12, December 2016
Morphology-Theory and Applications, vol.1, no.1, pp.136- [114] R. Garnett, T. Huegerich, C. Chui, and W. He , โA 174, 2016. universal noise removal algorithm with an impulse [98] M. Welk, A. Kleefeld, and M Breuร, โNon-adaptive and detectorโ, IEEE Transactions on image processing, vol.14, amoeba quantile filters for colour imagesโ, International no.11, pp. 1747-1754, 2005. Symposium on Mathematical Morphology and Its [115] B. Smolka, and D. Kusnik, โRobust local similarity Applications to Signal and Image Processing, Springer filter for the reduction of mixed Gaussian and impulsive International Publishing, 2015. noise in color digital imagesโ, Signal, Image and Video [99] Irum, Isma, M. Sharif, M. Raza, and S. Mohsin, โA Processing, vol.9, no.1, pp. 49-56, 2015. Nonlinear Hybrid Filter for Salt & Pepper Noise Removal [116] E. Abreu, M. Lightstone, S. K. Mitra, and K. Arakawa, from Color Imagesโ, Journal of applied research and โA new efficient approach for the removal of impulse noise technology, vol.13, no.1, pp.79-85, 2015. from highly corrupted imagesโ, IEEE transactions on [100] Taลmaz, Haci, H. Demirel, and G. Anbarjafari, image processing, vol.5, no.6 , pp. 1012-1025, 1996. โSatellite image enhancement by using dual tree complex [117] M. S. Moore, M. Gabbouj, and S. K. Mitra, โVector wavelet transformโ, Denoising And Illumination SD-ROM filter for removal of impulse noise from color Enhancement, vol.978, no.1, pp. 4673-0056, IEEE, 2012. imagesโ, Eurasip Conference, ECMCS, 1999. [101] A. Boyat, and B. K. Joshi โImage denoising using [118] V. S. Bhadouria, D. Ghoshal, and A. H. Siddiqi โA wavelet transform and median filteringโ, Nirma University new approach for high density saturated impulse noise International Conference on Engineering (NUiCONE), removal using decision-based coupled window median IEEE, 2013. filterโ, Signal, Image and Video Processing, vol.8, no.1, [102] H. Dong, and F. Wang, โImage-Denoising Based on pp.71-84, 2014. Bior Wavelet Transform and Median Filterโ, Photonics and [119] C. Tomasi, and R. Manduchi, โBilateral filtering for Optoelectronics (SOPO), Symposium on IEEE, 2012. gray and color imagesโ, Proceedings of IEEE 6th [103] M. Gabhel, and A. Hiradhar, โComparative Analysis of International Conference on Computer Vision (ICCV), pp. Various Denoising Techniques for MRI Image Using 839โ846, 1998. Waveletโ, International Journal of Science and Research, [120] L. Jin, C. Xiong, and H. Liu, โImproved bilateral filter vol.3, no. 9, pp.1212-1216, 2014. for suppressing mixed noise in color imagesโ, Digital Signal [104] R. S. Shalini Hema, and M. S. Godwin Premi, Processing, vol. 22, no.6, pp. 903-912, 2012. โComparative analysis of various wavelets for denoising color imagesโ, ARPN Journal of Engineering and Applied About the Authors Sciences, vol.10, no. 9, 2015, pp.3862-3867. Roji Chanu did BE and M.Tech from MIT [105] A. J. Bardos, and S. J. Sangwine, โSelective vector Manipur and NIT Silchar both in median filtering of colour imagesโ, Image Processing and Electronics and Communication Its Applications, 1997. Engineering. Currently, she is an assistant [106] B. Smolka, K. N. Plataniotis, R. Lukac and A. N. professor at NIT Nagaland. She is pursuing Venetsanopoulos, โSimilarity based impulsive noise PhD in the field of Digital Image removal in color imagesโ, International conference on Processing. She takes interest in Signal Multimedia and Expo, IEEE, 2003. Processing, Image Processing etc. [107] B. Smolka, R. Lukac, A. Chydzinski, K. N. Plataniotis, and W. Wojciechowski, โFast adaptive similarity based impulsive noise reduction filterโ, Real-Time Imaging, vol. 9,
no.4, pp.261-276, 2003. Kh. Manglem Singh did BE, MTech, MS [108] M. Szczepanski, B. Smolka. K. N. Plataniotis, and A. and PhD in Electrical Engineering, Control N. Venetsanopoulos, โOn the geodesic paths approach to & Instrumentation, Software Systems and color image filteringโ, Signal Processing, vol. 83, no.6, Digital Image Processing from DEI, pp.1309-1342, 2003. Agra, Delhi University, BITS Pilani and [109] Z. Wang, and D. Zhang โRestoration of impulse noise IIT Guwahati respectively. Currently, he is corrupted images using long-range correlationโ, IEEE an Associate Professor at NIT Manipur. Signal Processing Letters, vol. 5, no.1, pp. 4-7, 1998. His areas of interest are Image processing, [110] D. Zhang, and Z. Wang, โImage information Digital Watermarking, Steganography, and restoration based on long-range correlationโ, IEEE Information Security etc. Transactions on Circuits and Systems for Video Technology, vol.12, no.5, pp.331-341, 2002. [111] T. Sun, and Y. Neuvo, โDetail-preserving median based filters in image processingโ, Pattern Recognition Letters, vol.15, no.4, pp. 341-347, 1994. [112] D. Zhang, and Z. Wang, โImpulse noise removal using polynomial approximationโ, Optical Engineering, vol.37, no.4, pp. 1275-1282, 1998. [113] L. Khriji, F. A. Cheikh, and M. Gabbouj, โHigh- resolution digital resampling using vector rational filtersโ, Optical Engineering, vol.38, no.5, pp.893-901, 1999.