基于二代Curvelet变换与近似度制约规则的遥感图像融合算法

为了提高遥感图像的融合质量,消除模糊与不连续效应,设计了基于二代Curvelet变换与近似度制约规则的遥感图像融合算法。通过HSV变换对多光谱图像进行分解,获取其亮度子图。引入二代Curvelet变换,对全色图像与亮度子图进行精细分解,求取图像的高频与低频系数。利用图像的对角信息改进空间频率模型,求取全色图像低频系数的注入因子,并将全色图像的低频系数注入到亮度子图对应的低频系数中,获取融合低频系数。利用结构相似度模型对不同高频系数进行近似度测量,建立近似度制约规则,获取融合高频系数。通过对融合系数实施Curvelet和HSV反变换,获取融合的遥感图像。实验结果表明,较已有的遥感图像融合方法而言,所提方法具有更好的融合质量,包含了更多的光谱与空间信息。

Remote sensing image fusion algorithm based on second generation Curvelet transform and approximation constraint rule

In order to improve the fusion quality of remote sensing images and eliminate the effects of blur and discontinuity,a remote sensing image fusion algorithm based on second generation curvelet transform and approximation constraint rule was designed. The multispectral image is decomposed by HSV transform to obtain its brightness sub-image. And the second-generation transform was introduced to decompose the panchromatic image and brightness sub-image for obtaining the high and low frequency coefficients of the image. The spatial frequency model was improved by using the diagonal information of the image,and the injection factors of the corresponding low frequency coefficients of the panchromatic image were obtained. The low frequency coefficients of the panchromatic image are injected into the corresponding low frequency coefficients of the luminance subimage to obtain the fusion low frequency coefficients. The approximation of different high frequency coefficients was measured by structural similarity model, and the approximation restriction rules were established to obtain the fused high frequency coefficients. The inverse curvelet transform and HSV transformation was used to process the fused coefficients to obtain the fused image. The experimental results show that this method has better fusion effect than the current remote sensing image fusion method,which contains much spectral and spatial information.