Research on elliptical gear parameter measurement technology based on machine vision

In response to the issues of complex instrument operation, high costs, lengthy measurement times, and susceptibility to subjective errors in the traditional mechanical measurement of elliptical gear geometric parameters, a non-contact measurement system based on machine vision for elliptical gear parameters is proposed. Firstly, the collected elliptical gear images undergo preprocessing, histogram equalization and median filtering are applied to enhance the images. Subsequently, Zernike moments are employed to extract sub-pixel edge data of the elliptical gears. For the extracted data, the Hough transform method is used to determine the center of the inner hole. Then, the least squares method is applied to fit the tooth top curve and tooth root curve, ultimately solving for parameters such as pitch. Experiments conducted using an elliptical gear component indicate that the measured values meet the required measurement accuracy, demonstrating the feasibility of the proposed measurement system. The experimental results reveal that this system boasts high measurement accuracy, fast speed, simple operation, and low cost, making it applicable to the geometric parameter measurement of other types of non-circular gears and exhibiting broad application prospects.