Fast Template Matching for Vision- Based Localization

Fast Template Matching for Vision- Based Localization

FAST TEMPLATE MATCHING FOR VISION- BASED LOCALIZATION by JASON HARPER Submitted in partial fulfillment of the requirements For the degree of Master of Science Thesis Advisor: Wyatt Newman Department of Electrical Engineering and Computer Science CASE WESTERN RESERVE UNIVERSITY May 2009 Table of Contents Table of Contents ................................................................................................................................. 1 List of Figures ....................................................................................................................................... 4 List of Tables ........................................................................................................................................ 7 1 Introduction .................................................................................................................................. 9 1.1 Localization .......................................................................................................................... 9 1.1.1 Global Visual Localization ....................................................................................... 12 1.1.2 Local Visual Localization ......................................................................................... 13 1.2 Template Matching ............................................................................................................ 14 1.2.1 Edge Matching ........................................................................................................... 14 1.2.2 Pattern-Based Template Matching ......................................................................... 14 1.2.3 Grid Template Matching .......................................................................................... 15 1.3 Combining Vision Data with a State Observer ............................................................. 16 1.4 Contributions ..................................................................................................................... 17 2 Problem Definition .................................................................................................................... 18 3 Platform Setup ............................................................................................................................ 21 3.1 The Robot Platform .......................................................................................................... 21 3.2 The Chosen Camera .......................................................................................................... 22 3.3 Camera Calibration ............................................................................................................ 23 3.4 Timing Calibration ............................................................................................................. 28 1 4 Fast Template Matching ........................................................................................................... 32 4.1 Overview ............................................................................................................................. 32 4.2 Camera Data Acquisition ................................................................................................. 33 4.3 Finding Tiles ....................................................................................................................... 33 4.4 Plan View to Hough Space Transforms ......................................................................... 38 4.5 Transforming Hough Space into Modulo Space .......................................................... 41 4.6 Normalizing the Transform ............................................................................................. 46 4.7 Pose Estimation from Mod-Space for Localization ..................................................... 51 4.8 Making Template Matching Fast ..................................................................................... 55 5 Integration with a Localization System ................................................................................... 58 5.1 Overview ............................................................................................................................. 58 5.2 Variance from Helper Functions ..................................................................................... 58 5.3 Credibility Functions ......................................................................................................... 60 5.4 Setting Credibility Thresholds ......................................................................................... 64 6 Results .......................................................................................................................................... 74 6.1 System Results .................................................................................................................... 74 6.2 Algorithm Runtime ........................................................................................................... 79 6.3 Robustness with Respect to Noise .................................................................................. 81 6.4 Resistance to Distracter Lines ......................................................................................... 82 6.5 Integration with Localization System ............................................................................. 84 2 7 Conclusions ................................................................................................................................. 89 8 Future Work ............................................................................................................................... 91 Appendix A ......................................................................................................................................... 93 Appendix B ......................................................................................................................................... 94 B.1 Canny Edge Detector ........................................................................................................ 94 B.2 Laplace Edge Detector ..................................................................................................... 94 B.3 Weighted Local Variance .................................................................................................. 94 Bibliography ........................................................................................................................................ 96 3 List of Figures Figure 1: ALEN, the robot used for algorithm testing. ................................................................ 21 Figure 2: Firefly MV from Point Gray Research (Firefly MV, 2008). ........................................ 22 Figure 3: Camera view of aligned grid points. ................................................................................ 25 Figure 4: Standard deviation of the i-components manually chosen for Cartesian space subset. ................................................................................................................................................... 26 Figure 5: Standard deviation of the j-component of the Cartesian space subset. ..................... 26 Figure 6: Cartesian space image transformed from camera space using the calibration lookup table. ..................................................................................................................................................... 28 Figure 7: Image of circuit used to provide a timestamp for timing calibration. ....................... 29 Figure 8: Example of binary timestamp 0b00100101101011. ..................................................... 30 Figure 9: Timing delay of camera to PSO. ..................................................................................... 31 Figure 10: Pipeline architecture diagram. ........................................................................................ 33 Figure 11: Edge detection of hallway image using the Sobel operator....................................... 34 Figure 12: Edge detection of hallway image using the Canny algorithm. .................................. 34 Figure 13: Edge detection of hallway image using the Laplace convolution kernel. ................ 35 Figure 14: Edge detection of hallway image using the weighted local variance method. ........ 35 Figure 15: Hough space transformed from output of Sobel Figure 13...................................... 40 Figure 16: Example of distracting lines. .......................................................................................... 42 Figure 17: Illustration of equations 8 and 9.................................................................................... 43 Figure 18: Sample sections of mod space. ...................................................................................... 46 Figure 19: Mod space with distracter lines. .................................................................................... 47 Figure 20: Mod space after normalization. ..................................................................................... 50 4 Figure 21: Process of projecting mod space and creating a combined confidence function. 52 Figure 22: Depiction of creating the x-confidence function. ....................................................... 52 Figure 23: Depiction of creating the y-confidence function. ....................................................... 53 Figure 24:

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    105 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us