Product packaging quality detection method and system based on deep learning

Abstract

The invention relates to the technical field of image detection, and discloses a product packaging quality detection method based on deep learning, and the method comprises the steps: obtaining a product appearance image, carrying out the preprocessing of image graying and gray stretching of the product appearance image, and obtaining a preprocessed product appearance image; performing noise reduction processing on the preprocessed product appearance image by using a noise reduction algorithm based on similar image blocks to obtain a denoised product appearance image; processing the denoised product appearance image by using an image reconstruction algorithm to obtain a super-resolution product image; performing feature extraction on the super-resolution product image by using a feature extraction algorithm based on the sparse dictionary to obtain packaging features in the product image; and detecting the packaging quality of the product by using a packaging defect detection algorithmbased on the spatial pyramid pool. The invention further provides a product packaging quality detection system based on deep learning. According to the method, image detection is realized.

Description

technical field [0001] The present invention relates to the technical field of image detection, in particular to a product packaging quality detection method and system based on deep learning. Background technique [0002] At present, the packaging quality inspection of industrial products is a traditional visual inspection algorithm that relies on manual feature extraction. However, due to the diversity of products and changing detection conditions, the algorithm needs to be changed frequently to meet the detection needs. However, deep learning can extract sample features by itself and reduce costs. Using deep learning to inspect the quality of product packaging has become a hot topic in current research. [0003] However, in the process of collecting data in industrial production, there are often problems with few defect categories and inconspicuous defect characteristics, which are contrary to the large demand for deep learning samples and obvious characteristics; at the...

AI Technical Summary

Problems solved by technology

[0003] However, in the process of collecting data in industrial production, there are often problems with few defect categories and inconspicuous defect characteristics, which are contrary to the large demand for deep learning samples and obvious characteristics; at the same time, fog, rain and snow in bad weather conditions will not be able to solve the problem. To avoid being captured by imaging equipment, the color and contrast of objects in product images will be affected by these weather factors, especially rain and snow will block part of the image content, resulting in irreversible changes in the information contained in the image. Degradation will not only affect the visual quality of the image, but also change the important information contained in the image, thereby reducing the accuracy of product packaging quality inspection

Images