Model training method, figure image completion method and apparatus, and electronic device

Abstract

The invention relates to the technical field of image processing, in particular to a model training method, a figure image completion method, a model training device, a figure image completion device,a medium and electronic equipment. The model training method comprises the following steps: respectively preprocessing a figure sample image according to a preset algorithm, a preset random templateand a preset model to obtain a target image and a to-be-processed image with a pixel loss region; inputting the preset random template and the to-be-processed image into a generative neural network togenerate a completed figure image and a completed edge image; inputting the completed figure image, the completed edge image, the figure sample image and the target image into a discrimination neuralnetwork to generate a discrimination result; and performing multiple rounds of adversarial training on the generative neural network and the discrimination neural network according to discriminationresults corresponding to the plurality of figure sample images. According to the technical scheme of the embodiment of the invention, the corresponding surface texture can be complemented according tothe postures of different figures, so that the difference between the complemented figure image and the figure sample image is reduced.

Description

technical field [0001] The present disclosure relates to the technical field of image processing, and in particular, to a model training method, a character image completion method, a model training device, a character image completion device, a computer-readable storage medium, and electronic equipment. Background technique [0002] In the process of storing, transcoding, and transmitting images, it often happens that there are pixel loss areas in the image. In order to ensure the quality of the image, developers often complete the image in the following two ways: one is to fill in the pixel loss area by matching and copying the background block; the other is to train the generated confrontation network to make the generated The device can generate a complete complementary image consistent with the original missing image, so as to realize the completion of the original image. [0003] However, when encountering complex character images, the above-mentioned first method can...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present disclosure is to provide a model training method, a character image completion method, a model training device, a character image completion device, a computer-readable storage medium, and an electronic device, thereby at least to a certain extent overcoming the problem of completing character images. When the character's surface texture cannot be completed, there is a large gap between the completed image and the original image before the deletion.

Images