基于自适应阈值“双K算法”的零件缺陷边缘检测

针对传统边缘检测方法难以实现边缘信息的准确检测问题,提出了一种零件缺陷边缘检测的新方法.首先对采集到的零件缺陷图像进行灰度化和Wiener滤波,以减少噪声等因素对后期检测的影响;然后,以kalman算法预估图像分割阈值作为Krisch算法的初始阈值;在此基础上,进行零件缺陷边缘检测,以提高零件缺陷检测的准确性.最后,利用MATLAB软件对零件缺陷图像进行仿真试验,验证边缘检测算法的检测效果.实验结果表明,推荐算法检测的平均准确率达到94.38%,能够有效实现零件缺陷边缘信息的检测.

Part Defect Edge Detection Based on Adaptive Threshold "Double K Algorithm"

Aiming at the problem that the traditional edge detection method is difficult to accurately detect the edge information, a new method for edge detection of part defects is proposed. Firstly, the collected part defect image is grayed and Wiener filtered to reduce noise and other factors for later detection. The influence of the image segmentation threshold is estimated by the kalman algorithm as the initial threshold of the Krisch algorithm. On this basis, the edge detection of the part defect is performed to improve the accuracy of the defect detection of the part. Finally, the MATLAB software is used to image the defect of the part. The simulation experiment verifies the detection effect of the edge detection algorithm. The experimental results show that the average accuracy of the recommended algorithm detection is 94.38%, which can effectively detect the edge information of the part defect.