Research & Implementation Of Traffic Sign Recognition Algorithm Based On Multi-cues Hybrid

The traffic sign is an important component of the traffic system, traffic signs are used to guide drivers for direction, and to warn them of any special road condition. A real-time and robust traffic sign recognition can support and disburden the driver, and thus, significantly increase driving safety and comfort. How to accurate the use of real-time computer vision system to identify the complex environment of traffic signs under the Intelligent Transportation System has become a research hotspot.As a result of a wide range of traffic signs and often complex environment in outdoor conditions, the detection and recognition are vulnerable to weather, light, deformation, tilt, faded, similar to the background, signs and other adhesion factors, both in performance and to ensure that recognition rate to meet the real-time, so the design of recognition algorithm is quite difficult. This article has improved and innovated the existing recognition algorithm by summing up large various research methods at home and abroad, thus we have designed and realized an algorithm based on muti-cues hybrid.In detection stage, the color and gradient cues are used to segment the interesting regions, the corner and geometrical cues are used to detect the signs. A pseudo RGB-HSI conversion method without the need of nonlinear transformation is presented for color extraction.In recognition stage, a coarse classification is performed using the corresponding relationship of color and shape, then the Support Vector Machines with Binary Tree Architecture is built to recognize each category of traffic sign. Furthermore, we present a finite-state machine to decide whether a traffic sign is really recognized by fusion multi-frame recognition results or not.Most existing traffic sign tracking approaches including Kalman filters and particle filter suffer from the same limitations that ego-motion must be taken into account; otherwise it will affect the tracking performance. So Lucas-Kanade feature tracker is introduced for traffic sign tracking.Experimental results in different conditions, including sunny, cloudy, and rainy weather demonstrates that most traffic signs can be correctly detected and recognized with a high accuracy. It is find a practical solution to solve the existing problems and difficulties of the traffic sign recognition.