Product defect detection method and device

Abstract

The present invention provides a product defect detection method and device, wherein the method includes the following steps: constructing a deformable convolution defect detection model; obtaining image data of the product to be detected; equalizing the image data to obtain sample data; Data training deformable convolution defect detection model; use the trained deformable convolution defect detection model to perform defect detection on the product to be inspected. The present invention can ensure the balance of sample data, thereby preventing over-training of the deformable convolution defect detection model. In addition, it can also enhance the generalization ability and robustness of the deformable convolution defect detection model, thereby improving the deformable convolution defect detection model. Cross-platform performance of convolutional defect detection models.

Description

technical field [0001] The invention relates to the technical field of defect detection, in particular to a product defect detection method and a product defect detection device. Background technique [0002] When deep learning is applied to the training process of mobile phone surface defect detection, it is necessary to consider not only the imbalance of different defect sample categories, but also the generalization ability of the model. Different mobile phone projects generally need to build a new detection model and re-train it, resulting in weak generalization ability of the model. In addition, different mobile phone surface textures and colors are different. How to build a model with strong generalization ability is facing a challenge. big challenge. Contents of the invention [0003] In order to solve the above technical problems, the present invention provides a product defect detection method, which can ensure the balance of sample data, thereby preventing the o...

AI Technical Summary

Problems solved by technology

[0003] In order to solve the above technical problems, the present invention provides a product defect detection method, which can ensure the balance of sample data, thereby preventing the over-training of the deformable convolution defect detection model, and in addition, it can also enhance the deformable convolution defect detection The generalization ability and robustness of the model can improve the cross-platform performance of the deformable convolution defect detection model

Images