Image processing method and device

Abstract

The embodiment of the invention provides an image processing method and device, and the method comprises the steps: obtaining a to-be-processed image; inputting the to-be-processed image into a first image processing model for image segmentation to obtain an initial segmentation result, a contour probability mask and hidden layer feature parameters; and adopting a second image processing model to intercept a target image from the to-be-processed image according to the contour probability mask, and performing fine trimming processing on the target image according to the initial segmentation result and the hidden layer feature parameters to obtain a target segmentation result. The first image processing model generates a contour probability mask to provide a reference for a subsequent refinement process, so that an image segmentation network structure only corrects an uncertain region in a segmentation result, a large amount of redundant calculation is reduced, the image segmentation speed is improved, and the segmentation effect is very fine.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, in particular to an image processing method and an image processing device. Background technique [0002] In some video call scenarios, the background of the characters in the video needs to be shielded, or the background needs to be replaced arbitrarily for entertainment purposes. For example, when users conduct video chats with strangers through the video call function, some privacy-conscious users Sometimes you don't want to expose too much environmental information at home. After all, safety is the first priority when facing strangers. Therefore, it is necessary to block or replace the background in the video screen. [0003] In the prior art, by segmenting the background of the video image, and further shielding or replacing the background, the difficult problem in the field of video image segmentation is how to achieve fast and accurate segmentation, which is often a...

AI Technical Summary

Problems solved by technology

[0005] In view of the problem that the above-mentioned image segmentation algorithm has rough edge contour segmentation on high-definition images, and cannot be segmented in real-time in high-definition video due to the limitation of segmentation speed, the embodiments of the present invention are proposed to provide a method to overcome the above-mentioned problems or at least partially solve the above-mentioned problems. An image processing method for the problem and a corresponding image processing device

Images