Industrial product irregular defect detection method based on deep learning

Abstract

An industrial product irregular defect detection method based on deep learning comprises the following steps: firstly, performing image enhancement processing on an acquired sample image to enable defects to be more obvious; then, on the basis of a convolutional neural network CNN, constructing a defect detection network basic model in combination with the SSD target recognition model, and reasonably designing model parameters; finally, reducing the number of prediction boxes by adopting a non-maximum suppression algorithm, expanding a data set by adopting a data enhancement operation, increasing the training amount of the network, effectively improving the network identification performance can be, and solving the detection problem of irregular defects. The method is superior to a traditional detection method in the aspects of irregular defect detection, interference defect detection, detection real-time performance and the like, and can meet the requirements of enterprises for visualdetection of general industrial products.

Description

technical field [0001] The invention relates to a method for detecting irregular defects of industrial products based on deep learning. [0002] technical background [0003] In industrial production, the quality problems of industrial products are mainly manifested in problems such as production defects, assembly defects, various surface defects, and products that do not match the design. These quality problems are affected by many factors, such as production equipment, operators, and processing technology. Wait. Among them, the surface defect of the product is the main manifestation of the quality defect of the industrial product. The traditional surface defect detection method is manual visual detection, that is, human visual recognition in a specific environment, but this detection method has many disadvantages, such as high labor intensity, low work efficiency, high cost, and easy Affected by the quality and experience of the inspectors, etc. [0004] Industrialized l...

AI Technical Summary

Problems solved by technology

However, it is unavoidable that machine vision inspection technology still has some defects, such as: the detection effect of products with irregular defects is not good; it is limited by the computing power of the computer, which requires high computing power; it has real-time problems

In short, the traditional manual inspection method and some visual inspection technologies based on machine vision inspection have shortcomings, and cannot meet the needs of the industrial product inspection market. There is an urgent need for a detection method that meets market demand.

Images