Image super-resolution and fusion method based on regional information enhancement and block self-attention

Abstract

The invention relates to an image super-resolution and fusion method based on regional information enhancement and block self-attention, and belongs to the technical field of digital image processing.The method comprises a source image super-resolution branch and a fusion super-resolution branch. In a source image super-resolution branch, a feature extraction block is iteratively used to extracta source image feature map, and dense connections are used to fully utilize previous and subsequent feature map information. The output of each feature extraction block passes through a region information enhancement block to explore the region where each object in the source image is located, and the information assists in fusing the super-resolution branches to accurately predict the fusion decision graph. In the fusion super-resolution branch, two source images are spliced together to be input, and a fusion block based on a block self-attention mechanism is iteratively used in combination with region-enhanced source image information input in the source image super-resolution branch so as to better distinguish a focusing region from a non-focusing region. The last use of sub-pixel convolution of each branch produces a super-resolution source image and a fused image.

Description

Technical field [0001] The invention relates to an image super-division and fusion method based on area information enhancement and block self-attention, and belongs to the technical field of image information processing. Background technique [0002] The purpose of image fusion is to merge the information of two or more source images captured by different cameras in the same scene into one image, and to ensure that the information of each source image can be retained. Image fusion is widely used in security surveillance images, medical images, and satellite remote sensing images. In recent years, many studies have achieved good fusion results, but the existing methods are usually based on high-resolution multi-focus source image data sets to fuse. However, the images obtained by real imaging systems are not necessarily high Resolution image. When fusing low-resolution source images, the fused image will also be low-resolution, even fuzzy and lacking detailed information, which...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to provide an image super-separation and fusion method based on region information enhancement and block self-attention to solve the problem of image fusion when the resolution of the source image is low, so as to improve the quality of the fusion result

Images