Crack Defect Detection Processing Algorithm and Method of MEMS Devices Based on Image Processing Technology

In order to solve the problem that the crack defects generated on the surface of MEMS devices are difficult to detect under high overload impact, this paper proposes a crack detection method based on attribute weighted naive Bayes improved OTSU algorithm. Based on the analysis of the surface defects in MEMS devices image, the edge information of the crack defect in the image is extracted by image processing such as image detail sharpening, grayscale processing, image enhancement and edge extraction based on Canny operator, and the pseudo crack in the image is removed by the least square method; the attribute weighted naive Bayes algorithm is introduced to improve the traditional OTSU image processing method, the crack defect detection results of the MEMS devices image are obtained, the crack defects are quantitatively characterized, and the length and width of the crack defects are calculated. Comparative experiments were conducted using multiple detection methods, the results showed that the crack detection method proposed in this paper can obtain the crack defect information of MEMS devices efficiently and accurately.