Research On Tire Sensory Measurement System Based On Image Processing

With the improvement of people’s living standard, the car ownership is increasing and cars have become an indispensable part of people’s life. In driving, safety is always the most important problem people concern. But in automotive safety hazards, security risk of tire accounts for a large proportion, in which tire blowout is the major reason to cause the accidents. Except for the poor condition of roads, the most important reasons of tire puncture are the abnormal tire pressure and the wear of tire.Tire Pressure Monitoring System (TPMS) is widely used to detect tire pressure real-timely and related products are used extensively at home and abroad, but the existing TPMS have many shortcomings like the complexity of installation, the instability of signal and the invalidation of monitoring etc. Although there are some relevant provisions for TPMS mandatory installation in some foreign countries, it is hard to realize the TPMS mandatory installation in our country and some other countries. Research on tire wear detection is relatively small as well as related practical products. So, studying how to detect tire pressure and tire wear fast and accurately not only has a very important scientific significance, but also has very broad application prospects.This paper precisely researches the tire pressure, overload and wear detection algorithm based on image processing technology, and designs the hardware and software of the automotive tire sensory measurement system which has a TMS320DM642DSP as the core processor. It can detect the tire pressure、tire overload and degree of tire wear at the same time and be networked. It is a low-cost, high efficiency solution for tire sensory measurement and intelligent transportation. The main contributions of this paper are:(1) Presented and analyzed the mathematical morphology based tire positioning algorithm, and successfully extracted the tire part from the tire image;(2) Presented and analyzed three kinds of tire pressure and overload detection algorithms:the Tangent Method, the Ratio Method and the Euclidean Distance Method. Any one of the three methods has its own advantages and disadvantages, it can achieve a better detection results to use the three methods unitedly;(3) Had a research on the relationship between texture features and tire wear, mainly used the tire tread image’s Tamura texture features and GLCM-based texture features to analyze the normal and abnormal wear of tire.(4) Designed the hardware and software implementations of the tire sensory measurement system, designed and made a PCB board around DM642, given the design of hardware circuit and software flow chart of the major modules of the system and the software flow chart of the PC client.