{"id":3005,"date":"2020-02-24T09:45:06","date_gmt":"2020-02-24T09:45:06","guid":{"rendered":"http:\/\/www.sipotek.net\/?p=3005"},"modified":"2020-02-24T09:45:06","modified_gmt":"2020-02-24T09:45:06","slug":"can-ai-vision-inspection-sytem-inspect-cracks-on-product-surface","status":"publish","type":"post","link":"https:\/\/www.sipotek.net\/can-ai-vision-inspection-sytem-inspect-cracks-on-product-surface\/","title":{"rendered":"Can AI Vision Inspection Sytem Inspect Cracks on Product Surface?"},"content":{"rendered":"
Automated optical vision inspection system<\/a> is vision-based with image processing that discovers the defects by image analysis and vision eyes. Cracks ,damages, scratches are normal surface problems in quality inspection. Cracks inspection using non-destructive and non-contact methods are widely used in the industry. From small electronic parts to bulky wood, ceramics and wool, vision inspection method is cost effective, with high re-productivity<\/a> and reliability.<\/p>\n <\/p>\n For crack detection, a manufacturer might wish to tune the algorithm to detect cracks of specific length, shape and depth. In this case, 3D analysis is required to determine crack depth. The first step in the procedure is the actual detection of the cracks. With greyscale analysis, algorithms based on morphological segmentation are a simple choice used to detect an initial crack region. Characterization of the suspected crack in 3-dimensional is then performed in order to reduce false positives. Segmentation and extraction of the surface shape of the suspected crack allows to discern between true and false positive detection.<\/p>\n