Research on Color Correction Processing of Multi-Hyperspectral Remote Sensing Images Based on FCM Algorithm and Wallis Filtering

Abstract

This study proposes an improved color correction algorithm based on Fuzzy c-means (FCM) clustering algorithm and Wallis filtering to address the problem of spectral drift in image stitching. The proposed method can be applied to general multi-hyperspectral remote sensing images and focuses on color correction processing of images with and without overlapping areas. To ensure color correction, the color gamut is transformed using the color gamut transformation method, and an improved color matching algorithm based on Wallis filtering is used. For non-overlapping areas or hyperspectral remote sensing images, a feature class matching relationship is established using the class matching as the overlapping area for color correction processing, inspired by the application of FCM clustering in image segmentation. The experimental results show that the improvement of the histogram matching algorithm using Wallis filtering achieves a performance improvement of about 9.3%, and this performance is indexed by the spectral distortion. Also, the average gradient value of the image to be homogenized was significantly improved in the image color correction task without overlapping regions, where the false color image improvement was about 27.2%. The proposed approach is time-efficient and robust, while ensuring the quality of color correction processing.