Mobile phone screen defect segmentation method, device and equipment based on converged network

Abstract

The invention belongs to the field of machine vision and defect detection, and particularly relates to a mobile phone screen defect segmentation method, device and equipment based on a converged network. The method comprises the following steps of acquiring mobile phone screen images including a defect image and a defect-free image, training a pre-established defect detection network by using thedefect image and using a transfer learning method, and obtaining a defect candidate box corresponding to the defect image, training a pre-established image reconstruction network by using the defect-free image, and recovering a background reconstruction image, performing difference operation on the defect image and the background reconstruction image, and obtaining a defect segmentation image by adopting a threshold segmentation mode, utilizing the position coordinates of the corresponding defect candidate boxes on the defect segmentation image to extract the corresponding defect parts of thedefect segmentation image under the position coordinates, and obtaining a final defect segmentation result. According to the method, the defect detection network and the image reconstruction network are combined, so that not only can a small defect target be detected, but also the defect image can be accurately segmented.

Description

technical field [0001] The invention belongs to the fields of machine vision and defect detection, and in particular relates to a mobile phone screen defect segmentation method, device and equipment based on a fusion network. Background technique [0002] As the first window of human-computer interaction, the mobile phone screen is used to display images and colors. Currently, mobile phone screens mainly use Thin-Film-Transistor Liquid-Crystal Display (TFT-LCD) and Organic Light-Emitting Diode (OLED). The structure of TFT-LCD is to place a liquid crystal cell between two parallel glass substrates, set a TFT (thin film transistor) on the lower substrate glass, and set a color filter on the upper substrate glass, and control the liquid crystal by changing the signal and voltage on the TFT Molecular rotation direction, so as to achieve the control of each pixel point polarized light exit or not to achieve the purpose of display. OLED does not need a backlight, and uses a very...

AI Technical Summary

Problems solved by technology

Among them, the semantic segmentation network needs to build an image dataset with annotation information and train it in a pre-built network model, but there are problems with few defect data and small defect targets, which may easily lead to missed detection and low segmentation accuracy

In addition, using the defect-free image to construct the image reconstruction network, and then segmenting the defect, this method has low data cost and better segmentation effect, but the segmentation result is affected by the reconstructed image to a certain extent, and it is easy to cause defect errors. check

[0006] When the semantic segmentation network constructs the training data set, the defect target is small and unclear, which makes the labeling cost of image information too high, and there is also the problem of low segmentation accuracy

When performing defect segmentation through the image reconstruction network, the network reconstructs the background of the entire image. During defect segmentation, affected by the reconstruction effect, it is easy to segment false defects, resulting in false detection of defects.

Therefore, the existing defect segmentation methods cannot achieve accurate defect segmentation results

Images