Fabric defect detection method and system based on over-complete convolutional neural network

Abstract

The invention discloses a fabric defect detection method and system based on an over-complete convolutional neural network, and relates to the technical field of fabric surface defect detection, and the method comprises the following steps: an image acquisition step: employing an image acquisition device to collect a fabric image; an image preprocessing step: preprocessing the fabric image to obtain a preprocessed fabric image; a detection step: inputting the preprocessed fabric image into a trained over-complete convolutional neural network model, and outputting a detection result; and a visualization and storage step: visualizing and storing the detection result. The convolutional neural network model structure adopted by the invention is composed of an over-complete branch and an under-complete branch, and the recognition of small defects and the refinement of boundaries are improved while the precision is ensured; the end-to-end detection mode of semantic segmentation is adopted to realize the identification of the defect part of the fabric, and the detection result is more beneficial to the subsequent extraction of the quantitative feature of the defect.

Description

technical field [0001] The invention relates to the technical field of fabric surface defect detection, in particular to a fabric defect detection method and system based on an overcomplete convolutional neural network. Background technique [0002] With the rapid development of the textile industry, people's control over the quality of fabrics is becoming more and more stringent. In the textile industry, various unfavorable factors such as human error, machine failure, and yarn breakage can easily cause fabric defects and affect product quality. It will cause huge economic losses to the enterprise, so fabric defect detection is one of the important links of quality control. [0003] With the development of computer vision technology and the improvement of GPU computing power, deep learning technology has developed rapidly. The ability of deep learning algorithms to transform images into complex, abstract feature representations by combining low-level features overcomes the...

AI Technical Summary

Problems solved by technology

As the depth of the network increases, although the model can extract more advanced and abstract features, the ensuing problem of resolution reduction caused by deepening the network makes the model unable to identify small defects or fine details. Boundaries of chemical defects

Images