Diaphragm flaw detection method based on reverse bottleneck structure deep convolutional network

Abstract

The invention discloses a diaphragm flaw detection method based on a reverse bottleneck structure deep convolutional network, and the method achieves the detection and calibration of flaws in a diaphragm through employing the reverse bottleneck structure deep convolutional network. The method comprises the steps of image acquisition, image segmentation, data annotation, network training, flaw detection, image splicing and the like. According to the method provided by the invention, the effectiveness of the deep convolutional network for image feature extraction is fully utilized, and the number of parameters can be greatly reduced by the reverse bottleneck structure under the condition that the detection precision is ensured, so that the purpose of quickly and accurately detecting the defect points in the diaphragm is achieved.

Description

technical field [0001] The invention relates to the fields of deep learning and computer vision, and specifically designs a membrane defect detection method based on a reverse bottleneck structure deep convolution network. Background technique [0002] With the rapid development of the electronic technology industry, various portable devices such as notebook computers, tablet computers, and mobile phones are widely used in daily life, and the display screen, the window of the human-computer interaction interface, is particularly important. Due to the advantages of high display quality, no electromagnetic radiation, wide viewing area, and low power consumption, LCD screens are used in almost all portable devices. The optical film on the LCD screen not only protects the LCD screen but also affects the clarity of the display. In the production process of the optical film, dust, scratches, uneven printing, etc. will cause defects in the optical film, and these defects directly ...

AI Technical Summary

Problems solved by technology

[0004] Purpose of the invention: In order to solve the problems of low recognition accuracy and slow speed of the traditional flaw detection method, the present invention provides a membrane flaw detection method based on the reverse bottleneck structure deep convolution network, by using the reverse bottleneck structure deep convolution network The product network realizes the detection and calibration of the defect points in the diaphragm

Images