A Multispectral Image Fusion Algorithm Based on the HIS Space and Improved Wavelet

Abstract

In order to optimize the multispectral image fusion, the paper proposes a fusion method based on HIS space and improved wavelet. The new method maps the multispectral image to HIS space and acquires components of intensity, hue and saturation. In addition, an improved wavelet algorithm is applied on the fusion of the intensity of multispectral image and the panchromatic image, the algorithm can both consider the fusion method of low-frequency and high-frequency sub-bands: On the low-frequency subband aspect, the paper introduces a border detector operator in order to collect the border information of images that can be the basis of selective fusion method; On the high-frequency sub-band aspect, it applies local-standard-deviation to be the basis of selective and weighted-averaging fusion method. Ultimately, the new HIS components obtain fusion result through reverse transformation. The experiment reveals that the result acquired by this method can improve on the quality of spectral information and spatial information.