The Acquisition of Tooth Profile Curve Based on Image Recognition

Abstract

Gear wear is a common failure mode of gears transmission. Their usual monitoring methods are influenced by the machine suffered the load, lubrication conditions and other factors. These factors affect the accuracy of monitoring. Therefore, use image recognition technology to overcome these deficiencies, research on diagnosis of gear wear. Studies obtained a better method of gray cutting which can enhance the gear profile, the tooth profile curve of geometric correction methods and formulas, tooth profile curve extraction of the optimal algorithm, and carried out the discrete studies of that extracted tooth profile curve, and made the research of extraction tooth profile curve discretization, as well as a comparative study with the curve which obtained in the analysis method. The results show that the method can obtain the satisfactory actual gear profile curve, obtained the comparison curve is more accurate, intuitive, easy to carry out comparative research, further through the image recognition to determine the degree of gear wear failure to provide possible.