基于小波变换的图像融合技术研究

基于小波变换的融合算法形成了多分辨率融合算法的一个新的有用框架。由于人类视觉只对亮度信号的局部对比度敏感,提出了小波对比度的概念。图像经小波多分辨率分解后,依照小波对比度,定义变换系数邻域活性和系数匹配程度测度,按照系数匹配程度加权系数形成融合图像。将小波多分辨率分解和人类视觉对局部对比度敏感的特性优雅地结合在一起。

Research on Image Fusion Technique Based on Wavelet Transform

The image fusion algorithms based on wavelet transform form a new scheme for fusing multi-resolution images. The concept of wavelet contrast was raised according to the fact that the human vision only be sensitive to local brightness contrast. The neighborhood activity with transform coefficient and the measures for coefficient matching degree were defined according to the wavelet contrast after an image was decomposed, the result fused image was formed according to the wavelet contrast combined with the weight coefficients that was calculated based on the coefficients of matching degree. This technique combines the image multi-revolution decomposition by wavelet with local sensitive characteristics of human vision.