Defect classification and recognition method and device, computer equipment and storage medium

Abstract

The invention relates to a defect classification and recognition method and device, computer equipment and a storage medium. The method comprises the steps of obtaining an original image; segmenting the original image into a plurality of sub-images according to a preset sliding window and a sliding step length, inputting each sub-image into a trained defect recognition model; and extracting imagefeatures of each sub-image through the trained deep convolutional network model, inputting the image features of each sub-image into the classification and recognition model to obtain a classificationand recognition result of defects of each sub-image, and determining a defect type of the original image according to the classification and recognition result of the defects of each sub-image. The image is segmented, the features of the segmented image are quickly and accurately extracted through the trained deep convolutional network model, the extracted features are classified and recognized through the classification and recognition model to obtain the corresponding classification and recognition result, and the recognition speed and recognition accuracy are improved.

Description

technical field [0001] The present application relates to the field of computer technology, and in particular to a defect classification and recognition method, device, computer equipment and storage medium. Background technique [0002] At present, the detection of surface defects of materials mainly relies on manual detection, through on-site observation of processed products by workers, and evaluation of defects to determine whether the product is qualified. Since it is difficult to achieve a unified subjective standard for inspectors, it is difficult to meet the real-time and accuracy of inspection requirements. As the speed of industrial production becomes faster and faster, the cost of manual input for detection is increasing, the real-time performance is seriously affected, and the false detection rate increases accordingly. In order to solve the above technical problems, technicians have developed corresponding automatic detection methods and devices, mainly using i...

AI Technical Summary

Problems solved by technology

Due to the various forms of material surface defects and the large differences in scale and contrast, the existing detection methods are difficult to meet the efficient and accurate needs of industrial production

Images