LIU Bingxiang, HU Shiguo
(School of Information Engineering, Jingdezhen Ceramic Institute, Jingdezhen 333403, Jiangxi, China)

Abstract: Edge detection algorithm is an important part in image processing. The algorithm has a direct impact on the image processing results. The commonly used detection algorithm does not often have an ideal effect. In order to meet the testing requirements and results, it is necessary to improve the algorithm. In this paper, the digital image edge detection algorithm is improved by using the clustering algorithm.The improved solution is then applied to the repair of ancient ceramic cracks. The eexperimental comparison of the improved algorithm with the original shows that the improved algorithm has an obviously better detection result.

Key words: edge detection; ceramic repair; clustering algorithm