| With the rapid development of China's economy, the number of motor vehicles are also constantly increasing, we have to admit that motor vehicles bring a lot of convenience to our lives, but by the speeding violations and other issues also brought a lot of traffic problems, For example, traffic accidents, traffic congestion, transport efficiency. Based on these issues, intelligent transportation systems came into being, traffic sign recognition is an important part of the whole system, and more and more people pay attention on it. The function of system is recognizing the traffic sign base on the theory of Computer Vision and Machine Learning, the purpose is help drivers to identify the traffic signs in the process of driving. The system serve as a reminder and warning role, thereby reducing the incidence of accidents and improve safety.Through the analysis of traffic signs in our country, and the results of existing research studies, carried out improvement and innovation, ASM model is proposed based on speed limit signs recognition system, according to the unique shape of traffic signs, colors and other information to identify, The System mainly consists of two parts:(1) In the detection stage: we use the SVF operator in the color segmentation, an improved method based on gradient direction angle of the circle detection, experimental results show that the method under natural conditions of the block for the detection of speed limit signs have better results.(2) In the classification stage: The ASM model boundary point matching method for character recognition. The ASM model is a statistical model, which allows model in a certain range of deformation, suitable for traffic signs in natural scenes deformation conditions. Combined with the characteristics of the video, reduce computation, to achieve real-time.Finally, we used a variety of video data in case of the corresponding test results show that this algorithm has good real-time, accuracy and robustness. |
PDF Full Text Request