The Research And Implemention Of Tyre Codes Recognition System Based On Fuzzy

The car and other vehicle's quantity increases increasingly along with thedevelopment of our country highway and traffic. More and more tyres need to be provided. The traffic accidents, which the cars are destroied and peopleare bumped to die, occur continually because of the problems of the tyres' quality. The tyre factories are making great efforts to standardize production and management in order to improve tyres' quality. Now they want to recogniseevery tyre codes automatically and register the codes in the course of production so as to track, investigate, analyse and process their tyres. So it is imperative under the situation that we design a set of online tyre codes recognition software system for the factories. The system will bring prominent social benefits and economic benefits.In the process of developing the online tyre codes recognition software system, one of the key techniques is to recognise characters in the tyre codes image. This paper studies the image processing theories, the pattern recognition theories and other correlative theories. The paper sums up the possible problems of tyre codes image processing, and analyses the factors of affecting recognition efficeency, presents a method of uniting the character outline features and casting shadow features to recognise codes character. This paper also improves standard template matching method to recognise tyre codes. These two kinds of character recogniton methods can increase the rate of the character recognition. I have developed a set of integrated tyre codes recognition software system by using Visual C++ based on fuzzy theories and image/picture processing techniques. This system mainly realizes some following functions: Collecting and Saving the digital codes image, imagepre processing, each character image separating and mul ti-pattern recogrii tion. The system is practical and reliable. The system provides effective solution on practical problems of tyre codes recognition in the course of tyre production. 11 also improves work efficiency and automatic production and management level.The research has finished under Prof. Wang Staging's guidance. In theresearch, we fi rstly need to decrease image yawp, enhance image, separate image, and unitary character image process. Then we begin to recognise the tyre codes based on fuzzy theories by using improvement template matching method and the method of character out]ine features and casting shadow features. The test result of the recognition system indicates that the arithmetic of uniting the character outline features and casting shadow features and the arithmetic of improvement template matching are feasible and innovative. At the same time, based on the fuzzy set, the paper offers a new way to study other codes image characters recognition such as car license plate recognition, handwritten postal character recognition, container codes recognition etc. It has some ("unctions of theories, techniques reference and practical guidance.