loading...
 This Article 
   
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
Statistical Edge Detection: Learning and Evaluating Edge Cues
January 2003 (vol. 25 no. 1)
pp. 57-74
ASCII Text
x 
 
Scott Konishi, Alan L. Yuille, James M. Coughlan, Song Chun Zhu, "Statistical Edge Detection: Learning and Evaluating Edge Cues," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 25, no. 1, pp. 57-74, January, 2003.
 
 
BibTex
x
 
@article{ 10.1109/TPAMI.2003.1159946,
author = {Scott Konishi and Alan L. Yuille and James M. Coughlan and Song Chun Zhu},
title = {Statistical Edge Detection: Learning and Evaluating Edge Cues},
journal ={IEEE Transactions on Pattern Analysis and Machine Intelligence},
volume = {25},
number = {1},
issn = {0162-8828},
year = {2003},
pages = {57-74},
doi = {http://doi.ieeecomputersociety.org/10.1109/TPAMI.2003.1159946},
publisher = {IEEE Computer Society},
address = {Los Alamitos, CA, USA},
}
 
 
RefWorks Procite/RefMan/Endnote
x 
 
TY - JOUR
JO - IEEE Transactions on Pattern Analysis and Machine Intelligence
TI - Statistical Edge Detection: Learning and Evaluating Edge Cues
IS - 1
SN - 0162-8828
SP57
EP74
EPD - 57-74
A1 - Scott Konishi,
A1 - Alan L. Yuille,
A1 - James M. Coughlan,
A1 - Song Chun Zhu,
PY - 2003
KW - Edge detection
KW - statistical learning
KW - performance analysis
KW - bayesian inference.
VL - 25
JA - IEEE Transactions on Pattern Analysis and Machine Intelligence
ER -
 

Abstract—We formulate edge detection as statistical inference. This statistical edge detection is data driven, unlike standard methods for edge detection which are model based. For any set of edge detection filters (implementing local edge cues), we use presegmented images to learn the probability distributions of filter responses conditioned on whether they are evaluated on or off an edge. Edge detection is formulated as a discrimination task specified by a likelihood ratio test on the filter responses. This approach emphasizes the necessity of modeling the image background (the off-edges). We represent the conditional probability distributions nonparametrically and illustrate them on two different data sets of 100 (Sowerby) and 50 (South Florida) images. Multiple edges cues, including chrominance and multiple-scale, are combined by using their joint distributions. Hence, this cue combination is optimal in the statistical sense. We evaluate the effectiveness of different visual cues using the Chernoff information and Receiver Operator Characteristic (ROC) curves. This shows that our approach gives quantitatively better results than the Canny edge detector when the image background contains significant clutter. In addition, it enables us to determine the effectiveness of different edge cues and gives quantitative measures for the advantages of multilevel processing, for the use of chrominance, and for the relative effectiveness of different detectors. Furthermore, we show that we can learn these conditional distributions on one data set and adapt them to the other with only slight degradation of performance without knowing the ground truth on the second data set. This shows that our results are not purely domain specific. We apply the same approach to the spatial grouping of edge cues and obtain analogies to nonmaximal suppression and hysteresis.

[1] J. Atick and N. Redlich, "What Does the Retina Know about Natural Scene?" Neural Computation, Vol. 4, No. 2, 1992, pp. 196-210.[2] R. Balboa, PhD Thesis. Dept. of Computer Science. Univ. of Alicante, Spain, 1997.[3] R. Balboa and N.M. Grzywacz, “The Minimal Local-Asperity Hypothesis of Early Retinal Lateral Inhibition,” Neural Computation, vol. 12, pp. 1485-1517, 2000.[4] R. Balboa and N.M. Grzywacz, “The Distribution of Contrasts and its Relationship with Occlusions in Natural Images,” Vision Research, 2000.[5] Active Vision. A. Blake and A.L. Yuille, eds., Boston: MIT Press, 1992.[6] Empirical Evaluation Techniques in Computer Vision. K.W. Bowyer and J. Phillips, eds., IEEE Computer Society Press, 1998.[7] K. Bowyer, C. Kranenburg, and S. Dougherty, Edge Detector Evaluation Using Empirical ROC Curves Proc. Conf. Computer Vision and Pattern Recognition, 1999.[8] K.W. Bowyer, C. Kranenburg, and S. Dougherty, “Edge Detector Evaluation Using Empirical ROC Curves,” Computer Vision and Image Understanding, vol. 84, no. 10, pp 77-103, 2001.[9] J. Canny, “A Computational Approach to Edge Detection,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 8, no. 6, pp. 679-698, June 1986.[10] J. Coughlan, D. Snow, C. English, and A.L. Yuille, “Efficient Optimization of a Deformable Template Using Dynamic Programming” Proc. Computer Vision and Pattern Recognition (CVPR '98), 1998.[11] T.M. Cover and J.A. Thomas, Elements of Information Theory. John Wiley&Sons, 1991.[12] D.J. Field, “Relations between the Statistics and Natural Images and the Responses Properties of Cortical Cells” J. Optical Soc. Am., vol. A, no. 4, pp. 2379-2394, 1987.[13] S. Geman and D. Geman, “Stochastic Relaxation, Gibbs Distributions and the Bayesian Restoration of Images” IEEI Trans. Pattern Analysis and Machine Intelligence, vol. 6, pp 721-741, 1984.[14] D. Geman and B. Jedynak, An Active Testing Model for Tracking Roads in Satellite Images IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 18, no. 1, pp. 1-14, Jan. 1996.[15] D.M. Green and J.A. Swets, Signal Detection Theory and Psychophysics, second ed. Los Altos, Calif: Peninsula Publishing, 1988.[16] N.M. Grzywacz and R.M. Balboa, “A Bayesian Theory of Adaptation and Its Application to the Retina,” Neural Computation, vol. 14, pp. 543-559, 2002.[17] F. Heitger, L. Rosenthaler, R. von der Heydt, E. Peterhans, and D. Kubler, “Simulation of Neural Contour Mechanisms: From Simple to End-Stopped Cells,” Vision Research, vol. 32, pp. 963-981, 1992.[18] M. Kass, A. Witkin, and D. Terzopoulos, ”Snakes: Active Contour Models,” Proc. First Int'l Conf. Computer Vision, pp. 259-268, 1987.[19] Perception as Bayesian Inference. D.C. Knill and W. Richards eds., Cambridge Univ. Press, 1996.[20] S. Konishi, A.L. Yuille, J. Coughlan, and S.C. Zhu, Fundamental Bounds on Edge Detection: An Information Theoretic Evaluation of Different Edge Cues Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp. 573-579, 1999.[21] S. Konishi and A.L. Yuille, “Bayesian Segmentation of Scenes Using Domain Specific Knowledge” Proc. Computer Vision and Pattern Recognition, 2000.[22] S. Konishi, A.L. Yuille, and J.M. Coughlan, “A Statistical Approach to MultiScale Edge Detection,” Proc. Workshop Generative-Model-Based Vision: GMBV '02, 2002.[23] S. Konishi PhD Thesis, Dept. of Biophysics, Univ. of California at Berkeley, 2002.[24] A.B. Lee, J.G. Huang, and D.B. Mumford, “Random Collage Model for Natural Images” Int'l J. Computer Vision, Oct. 2000.[25] D. Marr, Vision. San Francisco: W.H. Freeman and Co., 1982.[26] M. Nitzberg, D. Mumford, and T. Shiota, “Filtering, Segmentation and Depth Filtering, Segmentation and Depth,” Lecture Notes in Computer Science, vol. 662, 1993.[27] J. Peng and B. Bhanu, “Closed Loop Object Recognition Using Reinforcement Learning,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 20, no. 2, pp. 139-154, Feb. 1998.[28] P. Perona and J. Malik, "Detecting and Localizing Edges Composed of Steps, Peaks and Roofs," Proc. Third Int'l Conf. Computer Vision,Osaka, Japan, pp. 52-57, 1990.[29] B.D. Ripley, Pattern Recognition and Neural Networks. Cambridge Univ. Press, 1996.[30] D.L. Ruderman and W. Bialek, “Statistics of Natural Images: Scaling in the Woods” Physics Review Letter, vol. 73, pp. 814-817, 1994.[31] M.C. Shin, D. Goldof, and K.W. Bowyer, “Comparison of Edge Detectors Using an Object Recognition Task,” Proc. Computer Vision and Pattern Recognition Conf., pp 360-365, 1999.[32] H. Sidenblath, “Probabilistic Tracking and Reconstruction of 3D Human Motion in Monocular Video Sequences,” PhD Thesis, Royal Inst. of Technology, Stockholm, 2001.[33] J. Sullivan, A. Blake, M. Isard, and J. MacCormick, “Object Localization by Bayesian Correlation” Proc. Int'l Conf. Computer Vision, pp. 1068-1075, 1999.[34] V.N. Vapnik, Statistical Learning Theory, John Wiley&Sons, 1998.[35] M.J. Wainwright and E.P. Simoncelli, “Scale Mixtures of Gaussian and the Statistics of Natural Images,” Proc. Neural Information Processing Systems, pp. 855-861, 2000.[36] A.L. Yuille and J.M. Coughlan, “Fundamental Limits of Bayesian Inference: Order Parameters and Phase Transitions for Road Tracking,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 22, no. 2, pp. 160-173, Feb. 2000.[37] A.L. Yuille, J.M. Coughlan, and S.C. Zhu, “A Unified Framework for Performance Analysis of Bayesian Inference,” Proc. The Int'l Soc. for Optical Eng. (SPIE), Apr. 2000.[38] S.C. Zhu and A. Yuille, “Region Competition: Unifying Snakes, Region Growing and Bayes/MDL for Multiband Image Segmentation,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 18, pp. 884-900, 1996.[39] S.C. Zhu and D. Mumford, “Prior Learning and Gibbs Reaction-Diffusion,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 19, no. 11, Nov. 1997.[40] S.C. Zhu, Y.N. Wu, and D. Mumford, “Minimax Entropy Principle and Its Application to Texture Modeling,” Neural Computation, vol. 9, no 8, Nov. 1997.

Index Terms:
Edge detection, statistical learning, performance analysis, bayesian inference.
Citation:
Scott Konishi, Alan L. Yuille, James M. Coughlan, Song Chun Zhu, "Statistical Edge Detection: Learning and Evaluating Edge Cues," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 25, no. 1, pp. 57-74, Jan. 2003, doi:10.1109/TPAMI.2003.1159946
Usage of this product signifies your acceptance of the Terms of Use.