论文部分内容阅读
针对多光谱与全色图像的融合,提出了一种改进的基于àtrous小波变换的图像融合方法。首先对多光谱图像进行IHS(intensity-hue-saturation)变换,然后对PAN图像进行àtrous小波变换,再将PAN图像的细节分量添加到多光谱图像的强度分量中。为了控制添加的数量,从小波系数中提取出一个重要性测度函数,并根据阈值选择重要的PAN特征,用新的强度分量来取代多光谱图像的强度分量。最后再做IHS逆变换得到融合图像。实验结果表明,和IHS变换算法及常用的小波融合算法相比,本文提出的融合方法在光谱信息的保持与空间细节信息的增强两个方面的综合性能得到提高,具有更好的效果。
Aiming at the fusion of multispectral and panchromatic images, an improved image fusion method based on àtrous wavelet transform is proposed. Firstly, IHS (intensity-hue-saturation) transform of multispectral image is carried out. Then, the PAN image is transformed by àtrous wavelet transform, and then the detail component of PAN image is added to the intensity component of multispectral image. In order to control the number of additions, an importance measure function is extracted from the wavelet coefficients, and important PAN features are selected according to the threshold values, and the intensity components of the multispectral image are replaced by new intensity components. Finally IHS inverse transform to get the fusion image. The experimental results show that, compared with the IHS transform algorithm and the commonly used wavelet fusion algorithm, the fusion method proposed in this paper can improve the overall performance of both the maintenance of spectral information and the enhancement of spatial detail information, with better results.