Research On Traffic Sign Recognition Based On Support Vector Machine

As science and technology development and social progress, urban highwaycongestion, especially the growing phenomenon of frequent traffic accidents, traffic safetyhas become one of the main issues. So intelligent transportation system (ITS) emerges anddevelopes rapidly. ITS is a comprehensive discipline, which involves the technologies ofimage processing, pattern recognition, artificial intelligence, electronic technology, controltechnology and information engineering technology and so on.Traffic sign recognition system (TSR) as an important part of the intelligenttransportation systems has become a hot topic of research. Traffic sign recognition isgenerally divided into of two stages: detection and recognition .This article conducted ain-depth research on some key technologies of the two stages.Through the analysis of the characteristics of the traffic sign, traffic sign detectionuse the algorithm combination of the color feature and shape feature. Firstly, colorsegmentation of traffic sign used a color difference method based on the RGB color space.Secondly, we removed a lot of interference background through regional notation and theimage was processed with mathematical morphology, Thirdly, Canny operator on theimage for the edge of the image and the paper used the improved Hough transform forshape extraction. Last, traffic sign was gotten through skiagraphy.Based on the statistical learning theory and support vector machine, the paper presenta classifier based on support vector machine for traffic sign recognition. Momentinvariants were extracted as the database in order to training and classification of theclassifier. Then the system of the traffic sign recognition was completed. This methodoptimized the system flow, improved accuracy, while reducing the complexity of thesystem.