面向高保真复现的多光谱图像融合算法

Multispectral Image Fusion Algorithm for High-Fidelity Reproduction

多光谱图像融合是精确获取完整图像信息的重要技术手段,本研究面向多光谱图像高保真复现领域,基于IHS变换与小波变换两种算法,对多光谱图像与高分辨率图像进行融合,采用信息熵、标准差、平均梯度、光谱扭曲程度、感觉信噪比5个指标对图像融合效果进行评价。结果表明,无论是两幅遥感图像的融合,还是多光谱图像与高分辨率图像的融合,IHS变换与小波变换融合算法均有良好效果,可为图像信息高保真复现提供良好的信息源。

Multispectral image fusion is an important technical means to achieve complete and accurate image information acquisition.For the high-fidelity reproduction field of multispectral images and based on two algorithms of IHS transform and wavelet transform,the multispectral images and high-resolution images were fused.Besides it proposed the information entropy,standard deviation,average gradient,spectral distortion degree and sensory signalto-noise ratio to evaluate images fusion effect.The results showed whether it is the fusion of two remote sensing images or the fusion of multi-spectral image and high-resolution image,the image fusion algorithm that combined IHS transform and wavelet transform has a good effect and provides a good information source for high-fidelity reproduction of image information.