Abstract: In view of the characteristics that there are obvious bright defects and not obvious dark defects in the residual area of the object surface, a defect edge detection method for the residual area of the object surface is constructed based on the combination of the improved LOG operator and Otsu algorithm.Firstly, aiming at the problem that the traditional LOG operator is inaccurate in detecting the bright defect edges of the image, Wiener filtering, which can automatically adjust the blur factor according to image characteristics, is introduced to replace the Gaussian filtering in the traditional LOG operator, thereby improving the accuracy of bright defect detection.Secondly, aiming at the problem that the dark defect edge detection result is inaccurate, Otsu algorithm is used to analyze the gray histogram of dark defects in the image to automatically determine the threshold, thereby improving the accuracy of dark defect edge detection.Finally, pixel weighted average fusion algorithm is used to fuse the bright and dark defect edges detected from the object surface image, so as to realize the detection of residual defects on the object surface.Experimental results show that compared with using only improved LOG operator or Otsu algorithm, the number of defective pixels detected using the weighted fusion method is basically consistent with that in the original image, which enables more accurate detection of the bright and dark defect edges in the residual areas on the object’s surface in the image.