Part Coaxiality Detection Based on Polynomial Interpolation Subpixel Edge Detection Algorithm
Yao Xiao, Yi Luo, Yunpeng Xin, Xiaodong Wang · 2020 3rd World Conference on Mechanical Engineering and Intelligent Manufacturing (WCMEIM) · 2020
In order to meet the requirements of rapid detection of coaxiality and high precision in the assembly process of aerospace small parts, a subpixel edge detection algorithm based on polynomial interpolation is proposed. First, Gaussian filtering is used to preprocess the image, and then the gray image and the direction template are convolved to obtain the gradient image. Then the gradient image is binarized to obtain the edges of the parts, and finally the edge points are subjected to quadratic polynomial interpolation to obtain the subpixel edges. Improved least squares method is used to fit the circular edges of the parts and calculate the coaxiality. Through the edge extraction experiment of small parts, the accuracy of the coaxiality obtained by this algorithm (PIS) is better than that of algorithm based on canny operator. It is a method that can accurately calculate the coaxiality of small parts.