Registration method for hyperspectral images

Abstract

The invention provides a registration method for hyperspectral image. The method comprises steps: S1, band images photographed by a spectral camera are arranged according to a band sequence, and the known photographing sequence priori knowledge is used for dividing the band images into multiple groups; S2, the groups are ranked, and the groups are well ranked in a sequential end-to-end mode according to the photographing sequence; and S3, the group at the very beginning of the arrangement serves as a first group, according to a reference transformation mode, a first scene image is selected as a reference in the group, intra-group registration is completed, the last scene in the first group serves as a reference next to register a first scene image in a second group, the first scene image of the second group after registration serves as a reference to complete registration of the second group, and thus, inter-group and intra-group registration work can be completed sequentially. The image distortion problem after multiple times of sampling can be effectively solved, and an excellent registration precision result can be obtained.

Description

technical field [0001] The present invention relates to an image registration method, and more particularly, to a registration method for framed hyperspectral images. Background technique [0002] With the rapid development of remote sensing technology, people have higher and higher requirements for the accuracy of ground detection, and hyperspectral technology has the characteristics of map-spectrum integration, which makes it quickly become one of the research hotspots of quantitative remote sensing. Hyperspectral images can not only accurately reflect the unique spectral characteristics of ground objects, but also reflect the spectral differences between ground objects, organically combine the spatial distribution of ground objects with spectral reflectance characteristics, and break through the previous unilateral research status. It has been widely used and researched in many fields. Among them, the application in the field of agriculture has made great progress, and i...

AI Technical Summary

Problems solved by technology

However, due to its global nature, the ability to deal with complex geometric transformations between images is insufficient, and due to the need for a large number of optimization calculations, when the optimization algorithms are different, there will be certain differences in the results.

At the same time, since this algorithm is based on pixel-level registration, it is difficult for the registration accuracy to reach the sub-pixel level

Therefore, this method is more suitable for image registration with rigid transformation, and less suitable for images with complex geometric transformation

[0011] To sum up, in the prior art, the selection of the image matching strategy requires multiple matching processes, which leads to repeated sampling of the image, and a large loss of original information of the original image, and the influence of the error on the shooting sequence is also great. Without consideration, the matching accuracy of the obtained registration results will not be the optimal result

[0012] The existing technology will cause deformation or even mismatch when registering band images with weak textures. For registration based on feature points, this is mainly because traditional methods are difficult to achieve between band images with increased grayscale nonlinear transformation. Due to the extraction of effective homonymous points, the correlation of the descriptors generated based on gradient transformation is greatly affected when the image registration of these bands is performed, so that the extraction of homonymous points cannot be stably realized.

Images