Research Of Road Traffic Signs Recognition Technology Based On Embedded Platform

Under the development of social economy, the quantities and types of vehicles is also more and more, so traffic jams have occured frequently on the road, meanwhile traffic accident probability is also increasing, and therefore we should not only make a reasonable management of road traffic but also increase focus on traffic safety issues. Intelligent Transportation Systems(Intelligent Transportation System, ITS) is a system that its design use of information technology and control technology knowledge, and that can solve the urban traffic problems. It can be applied in the rational management of road traffic, which can improve the safety and efficiency of transport systems. Road traffic sign recognition system can automatically the details information of the traffic signs, and it is an important part of ITS, which plays important roles in the auxiliary driving, automatic driving, traffic signs maintenance and other aspects. At this stage, the machine vision technology is becoming matures, and embedded system become more and more perfect, which make more scholars begin to research road traffic sign recognition system.The paper focuses on the detection and recognition algorithms to study in traffic signs, and based on ARM microprocessor to achieve. In traffic sign detection, in order to quickly and accurately locate the traffic signs, first proposed a method based on the combination of HSV color space and morphological operations, and verified by the experimental method that is feasible and adds a certain accuracy shape characteristics, and then combines the shape of the traffic signs to pinpoint. In traffic sign recognition, presented a traffic sign recognition algorithms that based on HU invariant moments of geographic restriction. The algorithm reduces the noise of the image portion, meanwhile improves the computational efficiency by limiting the area of signs images.In system design, it has been build the Linux systems, cross-compiler environment and Linux kernel porting, and on the platform prepared traffic sign recognition software program by using OpenCV library, and achieves the traffic sign recognition system successfully.