| The purpose of Safety Driver Assistance System(SADS) is to help the driver be ableto drive safely and decrease traffic accidents’ happen. Traffic Sign Recognition(TSR) isabsolutely necessary part of SADS because it can transfers the traffic signs’ informationthrough recognizing traffic signs in real-time and accurately and could regulate driver’sactions to avoid speeding, which is very dangerous and could result in heavy trafficaccident. So presented a rapid and accurate target recognition system for traffic signs isvery important.This traffic sign recognition system for sign detection and recognition of two stagesof the key technologies of a more in-depth research, we design and implement usingsupport vector machine as neural network classifier traffic sign identification system.Commonly used in traffic sign recognition image preprocessing techniques aresummarized, compared and analyzed respectively. First, color segmentation in RGB spaceSecondly, after using the image of segmentation Canny operators are edge extractionThirdly, we cut the recovered images according to their own color and shape features.This paper use color features, the improved invariants moment of Hu and Zernikemoments to extract the invariant moment of the traffic signs; Finally, color and shapefeatures are respectively detected with SVM classification, so we can get theseclassification models respectively. After the model is tested by these test samples, theresults show that the feature fusion recognition method has a very high accuracy.The paper constructed a distributed fuzzy support Vector machine. The result showsthat our method speeds up training while possesses high precision of SVMs and can meetactual demands.The whole system consists of various sensors, data collection circuit, alarming circuitof sound and light, display circuit of LCD screen. The high-powered chip-S3C44BOX, asthe central processing unit, is used to to ensure that the system can achieve. |
PDF Full Text Request