Research On Image Feature Analysis System Of Automobile Tires Based On Multi-CCD

With the development of automobile manufacturing industry and the improvement ofpeople’s living standards, cars are increasingly chosen as the ideal means of transport forpeople to go out, and the construction of intercity highway and urban transportationinfrastructure has created conditions for the automobile transportation to become anindispensable mode of transport. Unfortunately, here come the security issues. From70%to80%traffic accidents occurred on the highway are caused due to tire problems, of whichnearly50%are tire blowout accidents. The main reasons of the tire blowout accidents arethe abnormal tire pressure. Once the tire surface is badly worn and the tire is not replaced intime, tire blowout is likely to occur in the case of high-speed or sudden brakes. Moreover,the difference of tire type and tire pressure of coaxial on both sides may lower the safety ofthe vehicle driving. Therefore, maintaining the standard tire pressure and timely detectionof abnormal tire pressure and tire tread wear is the key to prevent tire blowout.At present, tire pressure detection is mainly based on Tire Pressure Monitoring System(TPMS), and the detection of tire wear mainly relies on maintenance technicians todetermine. Although TPMS has been widely used at home and abroad, it will still take thesome time to popularize TPMS in all vehicles; tire tread wear can only be detected bymanual inspection in regular vehicle maintenance. These all above greatly reduce thepossibility to detect the problem tires early. Therefore, research on how to make real-timecomprehensive judgment on the tire pressure and tire tread wear has a very importantscientific significance and very broad application prospects.Based on the multi-CCD image processing, this article makes a comprehensive analysisof the tire pressure and tire tread wear of each tire of a car. Firstly, the target image wasobtained by each CCD camera through image processing. Then the article calculated therelated geometric parameters of the detected tire and proposed a proportional method tojudge whether the tires of coaxial on both sides were the same in specifications and size andwhether each tire’s tire pressure was normal. After that, the article proposed to use theregion extraction method and GLCM to do research on tire tread wear. Then the articlemade a comprehensive analysis of the tire pressure and tire tread wear of the tires of a vehicle. Finally, an embedded hardware platform and the corresponding algorithm aredesigned based on TMS320DM642DSP core processor.