Research On Detection Technology Of Tooth’s Surface Marks For Automobile Main Reducer

The main reducer is an important part of automobile transmission system and its function is to reduce speed and increase torque. The quality of main reducer has great effect on the performance of the transmission system. Cars are often assembled with hypoid gears with excellent performance during the assembling process. Due to repeated friction, gear contact marks will appear in the process of meshing. Observing the contact marks can not only know whether the main reducer is qualified, but also can reflect its level of accuracy. The size, location, gear ratio and tooth width of the contact marks can be regarded as an important parameter to evaluate the quality of the main reducer. Artificial visual inspection method is one of the traditional detection methods, which is easily influenced by subjective factors, its strong subjectivity and low accuracy and efficiency. Particular attention is paid to collect the image acquisition through the methods which combined with machine vision and the image processing technology, and then do the image edge detection and feature parameter extraction of the collected images, and the results would be one of the standards to judge the quality of the main reducer assembly.First of all, the characteristics and the usual influential factor of the gear contact are analyzed.On the basis of color image gray level change, the study focuses on the enhanced processing on the collected data of the tooth surface’s imprint image through the methods respectively as follows:the histogram equalization, piecewise linear gray level transformation, and the image smooth processing method. It has obtained the sound effect of the early results.Secondly, to the pretreatment of image edge detection, the anslysis was made with two kinds of edge detection technology, a total of five kinds of operator. First, it is the differential operator edge detection method, which mainly introduced the principle of edge detection, the process of the inference and the deficiencies of Roberts, Sobel and Prewitt operator. Second, the method of edge detection of linear filtering focuses on the LOG and Canny operator, which have a wide range of application in recent times, and analyzes its edge detection principle and methods. Comparing of the five kinds of operators handling results, concluded the most suitable operator for tooth’s surface mark detection.Finally, based on the previous image preprocessing, the characteristic parameters of tooth’s surface were extracted and analyzed. This dissertation calculated characteristic parameters, such as area, perimeter, center of mass position, gear ratio and tooth length ratio of mark image. According to the parameters, the author judges whether the quality of gear pair meets precision requirement.In this paper, it could be find out that not only combine the LabVIEW visual aide and MATLAB platform, but also set up the main reducer gear tooth’s surface impressions collection and detection system. This system includes tooth’s surface imprint acquisition, image enhancement, edge detection and feature parameter extraction and as such modules. By comparing the imprint of parameters and artificial identification acquired from the test system, the accuracy of the detection system was verified.