Research&Implementation Of Speed Limit Sign Recognition Algorithm

Over the last decade, with social progress, economic development and the popularity of cars, traffic accidents occur frequently, most of which must be due to human factors and related to speeding vehicles. Nowadays, speed limit sign, as an important part of traffic signs, has a strong binding on driving behavior, and setting speed limit sign is the most popular control mode of vehicle speed around the world. But for various reasons, it’s difficult for drivers in the driving process to find the speed limit sign as early as possible. Therefore, the automatic detection and recognition of speed limit sign is of great significance for safe driving.As being in complex outdoor environmental conditions, speed limit sign’s detection and recognition are often vulnerable to the weather, light, tilt, bleaching, and similar to the background, furthermore, the performance of the algorithms needs to consider of not only recognition rate but also real-time. Therefore, it’s quite a difficult to design recognition algorithms. This article has improved the existing recognition algorithm by summing up large various of research methods at home and abroad, and has designed and implemented a new type of speed limit sign recognition algorithm.In the detection phase, according to the circle features of the objects from the paper studying, we detect the candidate area of speed limit sign in the gray image using rapid radial symmetry. The proposed algorithm can overcome the impact of light and weather changes, solve the problem caused by similar background and mark adhesion, avoid of the time-consuming Hough transform, and improve the detection accuracy and reliability of the algorithm.In the recognition phase, firstly, we obtain the character area of speed limit sign by the proposed method of Largest Containing Circle. Then identify the target through fuzzy template matching and Support Vector Machine with Complete Binary Tree Architecture multi-classifiers respectively. By comparison, Support Vector Machine with Complete Binary Tree Architecture based on Super-Sphere is better than the method based on fuzzy template in recognizing the objects. Finally, to increase the recognition rate, we proposed a multi-frame fusion method, which can determine whether a speed limit sign is really being recognized or not.Experimental results show that the method has a high detection and recognition rate with speed limit signs in car video and certain robustness to the complicated scenes, including lighting and weather changes. It is with simple procedure and runs near real-time. So it found a practical solution to solve the existing problems and difficulties of the speed limit sign recognition, and laid a good foundation for the following research on traffic signs recognition.