基于对偶树复小波—Curvelet变换的自适应多传感图像融合算法

Self-Adaptive Multi-sensor Image Fusion Using Dual-tree Complex Wavelet and Curvelet Transform

  • 摘要: 提出了基于对偶树复小波—Curvelet变换的自适应多传感图像融合新算法.算法将全色图像和多光谱图像进行对偶树复小波—Curvelet变换分解后,针对不同的频率域特点选择不同的融合规则.对低频系数选取区域能量的加权系数自适应融合规则,对高频系数特性选用了区域特征自适应的融合规则.最后通过重构得到融合图像.将其他的融合算法和本文所提算法进行对比,结果表明,基于对偶树复小波—Curvelet变换的区域特征自适应的图像融合算法是一种有效可行的图像融合算法.

     

    Abstract: A new self-adaptive multi-sensors image fusion algorithm based on dual-tree complex wavelet and Curverlet transform(DT-CWT) is proposed.In this algorithm,the panchromatic image and multi-spectral image are decomposed by DT-CWT,and different fusion rules are selected for different frequency characteristic.Regional energy weight coefficients self-adaptive fusion rule is selected for low-frequency coefficients,and regional feature self-adaptive fusion rule is selected for high frequency coefficients.Finally,the fusion image is obtained by reconstruction.Comparing other fusion algorithms with this algorithm,the results show the regional feature self-adaptive image fusion algorithm based on the DT-CWT is a feasible and effective image fusion algorithm.

     

/

返回文章
返回