Defect detection method and defect detection device in industrial detection

Abstract

The invention provides a defect detection method and a defect detection device in industrial detection. The method comprises the following steps: acquiring a to-be-detected workpiece picture; inputting the workpiece picture into a trained defect detection model to obtain a defect prediction result, wherein the defect detection model comprises an MASK-RCNN model and an FPN model, and the defect prediction result comprises a preselection box; performing image registration on the workpiece picture to obtain a registered front and back background segmentation image; screening a pre-selection frame of a defect prediction result according to the registered front and back background segmentation images; and obtaining a defect detection result according to the screened pre-selection box. According to the method, image registration and the FPN model are utilized, the area generated by the preselection box is limited, the model can concentrate on detection of the optical surface area, the detection speed of the model is increased, and the omission ratio is reduced; and the pre-selection frames are screened through image registration, defects of non-optical surfaces can be screened out, and therefore the over-detection rate of the model is reduced.

Description

technical field [0001] The invention relates to the technical field of detection, in particular to a defect detection method in industrial detection and a defect detection device in industrial detection. Background technique [0002] In the field of industrial quality inspection, the speed, over-inspection rate, and missed-inspection rate of the inspection model are three important indicators for evaluating the inspection model. [0003] In related technologies, during the detection process, the detection model will detect all regions in the picture indiscriminately, which will waste computing power on non-optical surfaces (regions that do not need to be detected, such as the background), and will also lead to a decrease in detection speed, and The detection of targets on non-optical surfaces will also lead to a decrease in the accuracy of the model and an increase in the over-detection rate. Contents of the invention [0004] In order to solve the above-mentioned technic...

AI Technical Summary

Problems solved by technology

[0003] In related technologies, during the detection process, the detection model will detect all regions in the picture indiscriminately, which will waste computing power on non-optical surfaces (regions that do not need to be detected, such as the background), and will also lead to a decrease in detection speed, and Target detection on non-optical surfaces will also lead to a decrease in the accuracy of the model and an increase in the over-detection rate

Images