Industrial product surface defect adaptive detection method based on deep learning model AGLNet

Abstract

The invention provides an industrial product surface defect adaptive detection method based on a deep learning model AGLNet. The method is characterized in that the method comprises the following steps: S1 carrying out the image collection of the surface defect of an industrial product on an assembly line, and obtaining a defect image; S2 carrying out manual labeling on the defect image to obtain label files of different types; S3 performing image enhancement operation on the defect image to obtain an enhanced image; S4 taking the label files and the corresponding enhanced image as a surface defect data set; S5 constructing a defect detector model based on a deep learning network, taking the surface defect data set as input, and training defect data in the manufacturing process based on the deep learning model AGLNet; S6 performing real-time defect detection on the products on the industrial production line through the trained AGLNet model, and obtaining the types and position information of the defects; and S7 sorting and counting the types and the position information of the defects, and analyzing the causes of the defects.

Description

technical field [0001] The invention belongs to the technical field of product defect detection, and relates to an adaptive detection method for surface defects of industrial products based on a deep learning model AGLNet. Background technique [0002] The development direction of my country's industrial manufacturing industry is towards high reliability, high precision, zero defect and intelligent high-speed development. High precision, high speed and high stability are an important part of the defect detection and identification process of industrial products. When there are defects on the surface of product parts, it is an important basis to judge whether a product meets the industrial quality requirements. In the traditional industry, the identification and detection of surface defects of product parts is mainly completed through manual quality inspection, but manual detection methods consume a lot of manpower, and have problems such as low efficiency, low precision and ...

AI Technical Summary

Problems solved by technology

However, this detection method still has the following deficiencies, which are difficult to meet the actual needs of industrial production: first, the manual feature extraction process is complex, and the feature information is difficult to contain all defect features; second, there is still the problem of too much human intervention in feature extraction , relying on artificially designed features, it is difficult to have good portability; at the same time, in a more complex detection environment, complex defect space agglomeration or multi-target detection environment, the detection accuracy based on machine vision is relatively poor, and Poor generalization ability; finally, when the detected product type changes, all machine vision-based algorithms and parameters need to be redesigned and developed

Images