The Application of Image Processing Software for Analysis of Roentgenograms
Total Page:16
File Type:pdf, Size:1020Kb
92 X Research and Teaching of Physics in the Context of University Education Nitra, June 5 and 6, 2007 THE APPLICATION OF IMAGE PROCESSING SOFTWARE FOR ANALYSIS OF ROENTGENOGRAMS Jan Sedláček Abstract Till this time the roentgenograms (also the electronic ones) are evaluated mostly visually. There are many possibilities for evaluation of digital roentgenograms. The aim of our effort is the betterment of the visual readability of roentgenograms for the accurate determination of the eventual seed damage. There are some possibilities to improve the gained electronic image with using PC. We can use either special PC programs or readily available software for image processing. One of the popular special PC programs is the software from the system of Lucia. It is used in life science, criminalistics, materials, or quality controls applications. The function of the “edges detection” is applied for the findings of the eventual seed damage. A very good job for image processing can made readily available PC programs. The freeware of Neat Image is one of the best programs for noise rejection. There is needed to sharpen the image by elevation of contrast and set-up of brightness past noise reduction. We can use Adobe Photoshop or XnView programs for that purpose. The excellent universal freeware for image processing is ImageJ. It uses the functions of median filter for noise reduction, contrast enhancing and edges detection. Keywords: visual readability, image processing software, improvement of images, roentgenogram, Lucia, Neat Image, Adobe Photoshop, XnView, ImageJ. Introduction The digital roentgenogram is made by special sensor, where the space arranged CCD matrix of elements makes the image, which is saved as a digital file for the next process in computer. The size of the CCD elements conforms to one pixel and the resolution depends on geometric ordering of particular elements. The resolution is about 11 lines per 1 mm. It comes up to probably 550 dpi (11 lines per 10 mm x 2 x 25.4 inch ≈ 550 dot per inch). In that way we are able to distinguish objects of size about 0.02 mm – 0.03 mm in gained roentgenogram - [7]. The software for enhancement of visual readability of roentgenograms We evaluate seed roentgenograms for the purpose to find the step of their damage. The roentgenograms are evaluated (also the electronic ones) mostly visually. So the resolution of roentgenograms is a very important thing. The electronic roentgenograms have ten times worse resolution than the classical ones; therefore we seek to better their visual readability by PC adjustment at least. There are a lot of computer programs for image analyzing and processing gained information. For adjustment of the electronic roentgenograms we can use either special PC programs or ordinary programs for the image processing. The program of Lucia One of the special PC programs, which are used at the physics department of the Technical faculty of the Czech University of Agriculture, is the software from the system of Lucia. Lucia is the image 93 X Research and Teaching of Physics in the Context of University Education Nitra, June 5 and 6, 2007 analysis software developed for capturing images, archiving and analysis. It is used in life science, criminalistics, materials, or quality controls applications – [5]. For the findings of the eventual seed damage the function of the “edges detection” is applied – see the following figures. Fig.1,2 show the roentgenogram of the bean seed, what is damaged by an external impact. The dark lines are cracks in the seed. Fig.1 – the original roentgenogram of Fig.2 – fig.1 after the application “edges the damaged bean seed detection” by Lucia The readily available software for image processing There were applied and used some methods of digital image processing, which improve the visual readability of the original roentgenogram. The goals of the process were: 1. To put down the noise what originated during digitalization and transfer of the image. (The noise displays above all as a more expressive granulity of the scan.) 2. The image focusing by elevation of contrast and set-up of brightness for better resolution of particulars. There were used the following computer programs for the fruition. Ad 1): The digital noise can be suppressed by PC adjustment. We can use some of casual graphic editor – often they have a special function for the noise reduction, but not always they work convincingly. However there are special programs made for that purpose only. They take off the noise more precise – it is their main advantage. Always that function is the question of the compromise – the noise has to be suppressed at full blast and at the same time the image cannot be smoothed unnaturally (the lost of sharpness happens.) One of that special programs is Neat Image, which is supposed one the best ones - [2]. Ad 2): There is needed the image focusing by elevation of contrast and set-up of brightness past the noise reduction. We can use Adobe Photoshop or XnView programs. They are suitable both for that purpose. Adobe Photoshop is default professional software for image adjustment. XnView is one of the best image browser and convertor with the possibility of the basic adjustments as the change of greatness, the modification of colour number, the preset of contrast and brightness, the application of filters etc. It manages four hundreds different kinds of file formats for reading and fifty formats for writing - [6]. Fig.3-5 show the enhancement of quality of roentgenogram of other damaged bean seed. 94 X Research and Teaching of Physics in the Context of University Education Nitra, June 5 and 6, 2007 Fig.5 – fig.4 after Fig.3 – the original Fig.4 – fig.3 after taking roentgenogram of off the noise by Neat focusing (elevation of contrast) by XnView the damaged bean Image The program of ImageJ The ImageJ is free, universal software, which disposes of many functions of earlier mentioned programs – reducing of noise, enhancing of contrast, edges detection etc. It can display, edit, analyze, process, save and print 8-bit, 16-bit and 32-bit images. It can read many image formats including TIFF, GIF, JPEG, BMP, DICOM, FITS and "raw". It supports "stacks", a series of images that share a single window. It is multithreaded, so time-consuming operations such as image file reading can be performed in parallel with other operations. That program can calculate area and pixel value statistics of user-defined selections. It can measure distances and angles. It can create density histograms and line profile plots. It supports standard image processing functions such as contrast manipulation, sharpening, smoothing, edge detection and median filtering - [4]. Fig.6-9 show some possibilities of adjustments of roentgenograms for enhancement of their quality. We use such functions as the “median filter” (reduces noise in the active image by replacing each pixel with the median of the neighbouring pixel values - [3]), the “enhance contrast“ by “saturated pixels” (the “saturated pixels” value determines the number of pixels in the image that are allowed to become saturated; increasing this value will increase contrast - [3]) and the “find edges” (it uses a Sobel edge detector to highlight sharp changes in intensity in the active image or selection - [3]). Fig.6 – the original roentgenogram of the Fig.7 – fig.6 after the application of damaged bean seed “find edges” 95 X Research and Teaching of Physics in the Context of University Education Nitra, June 5 and 6, 2007 Fig.8 – fig.6 after taking off the noise Fig.9 – fig.8 after the application of by median filter and enhance contrast “find edges” After juxtaposition fig.7 and 9 we can see that our adjustments really increased the visual readability of the roentgenogram. Conclusion The aim of our effort is the betterment of the visual readability of roentgenograms for the accurate determination of the eventual seed damage. We have some possibilities to improve the gained electronic image with using PC. We can use either special PC programs or readily available software for image processing. One of the special PC programs is the software from the system of Lucia. The function of the “edges detection” is applied for the findings of the eventual seed damage. Suitable readily available PC programs for image processing are freeware of Neat Image (for noise rejection), Adobe Photoshop or XnView (for focusing of the image). A very good universal freeware for image processing is ImageJ. It disposes of many functions of earlier mentioned programs – reducing of noise, enhancing of contrast, edges detection etc. References 1. Grodek J. – Pecen J. – Grundas S. 2006. Application of ImageJ to X-ray Image Analysis of Grain Damage. 5th International Workshop APLS 2006, Czech University of Agriculture, Technical Faculty, Department of Physics, Prague, p. 15-19. 2. http://digi.zive.cz/h/fotoskola/AR.asp?ARI=169 3. http://imagejdocu.tudor.lu/imagej-documentation-wiki/gui-commands/ 4. http://rsb.info.nih.gov/ij/docs/intro.html 5. http://www.lim.cz/index.php?lang=cs&&inc=nise 6. http://www.slunecnice.cz/product/XnView/ 7. Pecen J. 2001. Rozlišovací schopnost elektronických rentgenogramů. Sborník konf. „Výskum a vyučovanie na katedrách fyziky v kontexte univerzitného vzdělávánia“, SPU, Nitra, p. 11-18. Acknowledgments This article was written within the scope of the research project VZ č. MSM 6046070905. Contact Address: RNDr. Jan Sedláček, Ph.D., Czech University of Agriculture, Technical Faculty, Department of Physics, Kamýcká 129, CZ-165 21 Prague 6, Czech Republic, tel. +420 224383286, e- mail: [email protected] .