Affiliation:
1. School of Ceramic, Pingdingshan University, Pingdingshan, Henan 467000, China
Abstract
Ancient ceramics is an important carrier of concretization and artistic transformation of Traditional Chinese culture. It is also an extremely indispensable link for the world to understand traditional Chinese culture. It is of great significance to ensure the quality of ceramic products and improve the reliability of products by nondestructive testing of ceramic microdefect cracks. It is necessary to extract the microdefect crack area first and describe the characteristics of the ceramic crack image with the gradient weighting feature of the model to complete the nondestructive detection of the crack image. The traditional method sets the pixel point and brightness threshold according to the pixel value of the microdefect area but ignores the description of the weighted feature of the image and completes the nondestructive detection of the microdefect crack image of ceramic products. In this study, the improved algorithm of edge detection based on cluster analysis was applied to the ancient ceramic crack repair. First, cluster analysis is used to optimize the Sobel operator in edge detection. Then, the gray value distribution of edge detection map is changed by the Clustering algorithm. Finally,the experimental results show that the contour crack trace and edge direction of the improved edge detection map are obviously enhanced by 20%, which is beneficial to improve the accuracy of ancient ceramic crack repair.
Subject
Computer Science Applications,Software
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献