The Detection Of Surface Defect On Metallic Diaphragm

With the rapid development of computer vision and image processing technology,the scope of their applications are also increasingly widespread, and they have beenwidely used in the printing industry, aviation, agriculture, medicine and many otherfields. In order to detect the surface defects of the metal diaphragm due to production orartificial scratch and quantify the length and width, this paper takes the stainless steelwafer, the small circular ring and large circular ring as the research object, throughanalyzing and studying large numbers of diaphragm images, and then raises a treatmentresolution based on image processing technology to extract and quantify surface defectsof the metal diaphragm.First, in the image-taking process, the edge of the diaphragm generates apertureeasily which is the biggest interference factor while extracting the defects. The secondimportant interference factor is the bulk particles and slight scratches appearing on thesurface of image capturing device. The solution we use is to obtain the diaphragmregions, and then assign the background color to all non-diaphragm regions. Aiming atthis problem, this paper designs a new algorithm to detect and quantify circular regionsbased on dynamic scanning. This algorithm has high-speed and quasi-precisioncharacteristics. Secondly, we pretreat the diaphragm image. The process includessharpening, denoising, normalization, and dynamic threshold segmentation, after thisstep, defect images can be mainly divided. Then, we extract the target outline of theimage by using the subtraction method based on the mathematical morphology. Finally,we quantify the length and width of the outline, and remove some useless areas. Andthis paper presents a scratch screening and quantitative algorithm based on multi-regioncontour tracking. The experimental result shows that this algorithm can achieve a betterbalance in terms of speed and accuracy.This paper mainly studies the extraction and quantification computing process ofthe surface defects on metal diaphragm. The research achievement has important guiding significance for using image processing technique to detect the surface qualityof circular metal diaphragm.