Image fusion method for reducing mosaic image malposition

Abstract

The invention relates to the field of image processing and computer graphics, and aims to improve mosaic image quality, weaken malposition and splicing gaps and obtain a visually natural and coherent fusion image under a situation that a splicing parameter is constant. Therefore, the method adopts the technical scheme that the image fusion method for reducing mosaic image malposition uses the splicing parameter to process an input mosaic image, and the pixels, which have a coincident coordinate, of the processed image are processed. Specifically, the method comprises the following steps that: S1: solving the pixel luminance difference of the coincident area of two mosaic images; S2: solving safe edge points and a smooth area; S3: searching a fusion path; and S4: optimizing image fusion. The method is mainly applied to the occasions of image processing and computer graphic processing.

Description

technical field [0001] The invention relates to the fields of image processing and computer graphics, in particular to the field of optimized fusion effects based on optimized visual experience when performing image matching and splicing. Specifically, it involves an image fusion method that reduces the misalignment of stitched images. Background technique [0002] Image splicing and fusion technology is a common digital image processing technology, which is widely used in practical application fields such as automotive electronics, drones, military, and remote sensing. This technology is to stitch and fuse two or more images with overlapping areas into one image with a wide viewing angle. The steps of commonly used image mosaic methods are (1) extract key information such as feature points in the image; (2) match these key information in the two images; (3) use the matched key information to calculate the Stitching parameters. Stitching parameters include image rotation,...

AI Technical Summary

Problems solved by technology

[0003] However, in practical applications, the images used for stitching are often taken by different cameras at different angles, resulting in large rotation and stretching changes between the stitched images

Some current image stitching algorithms are sensitive to rotation and stretch changes, which leads to a certain deviation between the calculated stitching parameters and the real stitching parameters

When the image is stitched according to this stitching parameter, some misalignment will occur at the stitching point, which will cause the edge of the image content to break, etc.

At the same time, due to differences in the performance of different cameras and differences in shooting environments (such as lighting, etc.), there will be differences in brightness and other differences in the overlapping areas of the two stitched images, resulting in obvious stitching traces in the fused image.

thereby reducing the visual perception

Images