Image alignment method based on illuminance constraint and grid deformation

Abstract

The invention discloses an image alignment method based on illuminance constraint and grid deformation. The image alignment method is used for accurately aligning overlapped images. The method comprises the steps of initializing a color mapping model, jointly optimizing grid coordinates and the color mapping model, and optimizing and adjusting model parameters, estimating the model parameters in the grid deformation steadily; and finally, calculating a local homography transformation matrix in each grid according to the corresponding relationship between the coordinates of the grid points before and after deformation, and realizing accurate alignment between the images by utilizing an image deformation and texture mapping technology. According to the invention, on the basis of the traditional geometric constraint, the illumination constraint is further utilized, so that the alignment algorithm can process scenes with not rich textures. In addition, color mapping model expansion is carried out on a traditional image deformation framework, so that the grid deformation-based image alignment algorithm can realize accurate alignment between images with significant color differences.

Description

technical field [0001] The invention belongs to the technical field of image alignment, and in particular relates to an image alignment technical scheme and system based on illumination constraints and grid deformation. Background technique [0002] Image alignment refers to the process of aligning images with a certain degree of overlap in the overlapping area. When the images that need to be aligned come from different perspectives with a certain degree of overlap, image alignment technology can be used to generate panoramic images with a larger field of view range, so it has a wide range of applications in virtual reality and augmented reality. When the images to be aligned come from the sequence of images captured by the same camera at different times, the image alignment results reflect the corresponding relationship of the image content in the time series, which is especially useful in multimedia applications such as video stabilization and video denoising. important....

AI Technical Summary

Problems solved by technology

Commonly used constraints usually include geometric constraints such as matching points and matching line segments between images, but such constraints are likely to cause underconstraints when the scene lacks texture, so that model parameters cannot be effectively estimated

At the same time, the possible mismatch between feature points and line segments will seriously affect the final parameter estimation results, so that the images cannot be aligned accurately

Images