Identification of Degraded Traffic Sign Symbols Using Multi-class Support Vector Machines

We present a novel classification method for recognizing traffic sign symbols undergoing image degradations. In order to cope with the degradations, it is desirable to use combined blur-affine invariants (CBAIs) of traffic sign symbols as the feature vectors. Combined invariants allow to recognize objects in the degraded scene without any restoration. In this research, multi-class support vector machines (M-SVMs) is applied to traffic sign classification and compared with backpropagation (BP) algorithm, which has been commonly used in neural network. Experimental results indicate that M-SVMs algorithm is superior to BP algorithm both on the classification accuracy and generalization performance of the classifier.

[1]  Ulrich H.-G. Kreßel,et al.  Pairwise classification and support vector machines , 1999 .

[2]  Gérard Dreyfus,et al.  Single-layer learning revisited: a stepwise procedure for building and training a neural network , 1989, NATO Neurocomputing.

[3]  Jan Flusser,et al.  Combined blur and affine moment invariants and their use in pattern recognition , 2003, Pattern Recognit..

[4]  Jan Flusser,et al.  Pattern recognition by affine moment invariants , 1993, Pattern Recognit..

[5]  Vladimir Vapnik,et al.  The Nature of Statistical Learning , 1995 .

[6]  Miguel Ángel Sotelo,et al.  Fast traffic sign detection and recognition under changing lighting conditions , 2006, 2006 IEEE Intelligent Transportation Systems Conference.

[7]  Yoshiaki Shirai,et al.  An active vision system for real-time traffic sign recognition , 2000, ITSC2000. 2000 IEEE Intelligent Transportation Systems. Proceedings (Cat. No.00TH8493).

[8]  H. Fleyeh.  Shadow And Highlight Invariant Colour Segmentation Algorithm For Traffic Signs , 2006, 2006 IEEE Conference on Cybernetics and Intelligent Systems.

[9]  Dariu Gavrila,et al.  Multi-feature hierarchical template matching using distance transforms , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[10]  Luis Moreno,et al.  Road traffic sign detection and classification , 1997, IEEE Trans. Ind. Electron..

[11]  S. Lafuente-Arroyo,et al.  Traffic sign shape classification evaluation. Part II. FFT applied to the signature of blobs , 2005, IEEE Proceedings. Intelligent Vehicles Symposium, 2005..

[12]  S. Lafuente-Arroyo,et al.  Traffic sign shape classification evaluation I: SVM using distance to borders , 2005, IEEE Proceedings. Intelligent Vehicles Symposium, 2005..

[13]  Visvanathan Ramesh,et al.  A system for traffic sign detection, tracking, and recognition using color, shape, and motion information , 2005, IEEE Proceedings. Intelligent Vehicles Symposium, 2005..

[14]  Sei-Wang Chen,et al.  An automatic road sign recognition system based on a computational model of human recognition processing , 2004, Comput. Vis. Image Underst..

[15]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[16]  Chih-Jen Lin,et al.  A comparison of methods for multiclass support vector machines , 2002, IEEE Trans. Neural Networks.

[17]  Vladimir Vapnik,et al.  An overview of statistical learning theory , 1999, IEEE Trans. Neural Networks.

[18]  Ying Zhang,et al.  Determination of blur and affine combined invariants by normalization , 2002, Pattern Recognit..

[19]  Thomas H. Reiss,et al.  The revised Fundamental Theorem of Moment Invariants , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[20]  Vladimir Cherkassky,et al.  Vapnik-Chervonenkis (VC) learning theory and its applications , 1999 .

[21]  Jan Flusser,et al.  Degraded Image Analysis: An Invariant Approach , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[22]  Vladimir Cherkassky,et al.  Learning from Data: Concepts, Theory, and Methods , 1998 .