Sub-Pixel Edge Detection Algorithm Based on Improved Zernike Moments

In machine vision high-precision size detection, sub-pixel edge detection algorithm is often used, but the common sub-pixel detection algorithm has the disadvantages of low positioning accuracy and large computation. The model in the actual image is gradual edge instead of ideal step edge. In this case, a Zernike moment sub-pixel edge detection algorithm based on gradual edge model is proposed in this paper. Experimental results show that the traditional Zernike moment algorithm based on step edge model meets the requirements of noise immunity, high accuracy and strong real-time performance in machine vision measurement.