Linjun Zhao, Ruisha Zhu
Academic Journal of Computing & Information Science
: In this paper, aiming at the sub-pixel edge contour detection method, the sub-pixel edge contour extraction model and camera imaging geometric model are established step by step. Firstly, the improved anti-tangent function model based on canny operator is adopted. Firstly, the image is grayed out, then the edge of canny operator is detected by Gaussian filter, and a better method of edge detection after extended smoothing is proposed. Extract the sub-pixel edge outline of the target, then detect and fit lines, curves and circles, and mark them with different colors. Then for the correction of the image captured by the camera, this paper uses an image calibration model based on Zhang Zhengyou camera calibration method. The camera is calibrated by the calibration picture given, the internal and external parameters of the camera are obtained, and the unit matrix is calculated. Using the given definition of calibration board information, the number of pixels per millimeter pixel in the image is estimated. The research is helpful to improve the accuracy of sub-pixel edge contour detection and improve the image edge detection of industrial parts to a certain extent.