In the machine vision automatic inspection technology, the biggest technical adjustment is to detect the surface defects such as cracks, scratches, discoloration, ablation and so on. The difficulty lies in the irregular shape, low depth and light contrast, which is often disturbed by the natural texture or pattern of the product surface.
Therefore, the detection of surface defects requires high-precision lighting, camera resolution, the relative position between the detected parts and the camera, as well as complex machine vision automatic detection technology.
Illumination is an extremely important factor in the detection of surface defects of products, because surface defects usually have the characteristics of low contrast and can be separated from the background by using the correct illumination. For planar convex defects, you can use low angle lighting to "illuminate" the edges of convex parts.
Diffuse light can be used to detect smooth defects or surface defects. Cracks or scratches should cover at least three or more pixels in the captured image. In some cases, multiple high-resolution cameras or a single camera are used for multiple shots, or more commonly, a line scan camera from 1K to 16K is used to detect the entire surface of the part at a specified resolution to meet the user's accuracy requirements.
In order to detect the cracks and scratches of products by machine vision automatic detection technology, it is necessary to understand how the surface defects of products are generated, and clearly classify the possible shapes of these defects in the image. For example, when powder metal parts are thin, cracks will appear during sintering. Based on the experience knowledge, the machine vision automatic detection technology can be concentrated in the defect producing area, and the detection direction is clear, which is helpful to improve the reliability of defect detection.
In the process of manual inspection, the shape of defects can be predicted by intuition and experience. For example, scratches usually appear as a series of straight lines or slightly curved lines, which are often subconsciously connected into a line by a psychological Gestalt ability. In machine vision automatic detection technology, this "Gestalt" ability can be roughly expressed by visual calculus.
In the course of a difficult test, testers may have different opinions - which is a scratch? Or is it just a bright spot that happens to form a line? This shows that the reliability of defect detection needs to be considered. If the surface defects are clear and the defect components are clear, the consistency and reliability of the detection results of machine vision automatic detection technology are higher than that of artificial vision inspection.
In order to detect small defects on the product surface, it is very important to ensure that the part surface is always in the focus area of the camera. Visual automatic detection technology is used to accurately detect cracks and scratches on products. For more information about machine vision automatic detection technology, please visit the official website of Shandong Hongbao Automation Co., Ltd.