A novel fast subpixel edge location method based on Sobel-OFMM

Edge detection is a decisive step in vision measurement process. A novel efficient subpixel edge location algorithm based on Sobel-OFMM is presented in the paper. This algorithm firstly makes use of gradient operator, Sobel, to extract all possible edge points at pixel level and then utilizes subpixel-level edge detection operator, Orthogonal Fourier-Mellin Moment which uses lower radial orders and rotation invariance of moments to describe specific objects in the image. Experiment results show that this algorithm could achieve higher location accuracy and consume less time compared with Zernike and OFMM subpixel edge location operator. It can be concluded that this algorithm is efficient to implement subpixel-level edge detection and improve measuring accuracy in vision measuring technology.