Intelligent Mobile Terminal Based Traffic Sign Recognition

Nowadays, the number of cars is increasing and traffic accidents are frequent, Intelligent Transportation Systems attracted widespread attention. Among them, the traffic signs recognition system based ob machine vision is an important part of the Intelligent Transportation Systems, and has become a hot research topic. At the same time, intelligent mobile terminal rise gradually and develop rapidly, this brings new opportunities and greater development platform for traffic signs recognition system. Therefore, traffic signs recognition technology research based on intelligent mobile terminal have a widespread application prospects.In this paper, we begin our research with the android operating system as intelligent terminal platform, combines image processing, pattern recognition and other theories, processing and analysis the video of traffic information that extracted from the real car and real scene, develop the real-time speed limit signs recognition and warning function. The difficulty and the key is to take into account the real-time accuracy and adaptability to a variety of situation for recognition of speed limit signs. The main contents are as follows:1. After the crude extract of the target area, using the basic features of the image, such as: Color, shape, etc, Split extract the area of the speed limit signs. Realize fast positioning of the speed limit signs.2. Analysis the gray histogram of suspected targets, distinguish between speed limit signs and impurities; using a local iteration binarization algorithm for binary computing suspected target area, extract speed limit digital.3. Using the geometrical characteristics and contact between them of the speed limit digital, extract speed limit digital accurately; and then match with digital template, achieve recognition of speed limit digital.Finally, the algorithm is packaged into APK and test on the phone in the real road, the results show that the warning speed of the system has met real-time requirements; recognition accuracy has met expectations, adaptability to all kinds of weather, road and other external environment. It has verified the feasibility and reliability of the results of this study, has the value of in-depth development, and is a meaningful attempt of application of machine vision theory in Intelligent Transportation.