基于清晰度和非下采样多聚焦图像融合

Multi-focus Image Fusion Based on Definition and Nonsubsamping

根据多聚焦图像的特点提出一种基于清晰度的NSCT图像融合算法。在清晰的区域,低频系数和高频系数全部采用清晰区域的系数,而从清晰到模糊过渡的区域,低频系数则取区域方差值最大,高频子带系数取区域能量值最大。该算法与梯度金字塔算法、小波融合算法和Contourlet融合算法进行比较,实验结果证明该方法融合后的图像与源图像具有最小均方差。

The paper proposes an algorithm based on definition and NSCT in multi-focus image fusion. The coefficient of definition area is still adopted as fusion coefficient. In area from definition to blur, regional variance is adopted in low-frequency image, and maximum local energy is adopted in high-frequents subbands. Experimental results show that the proposed method has smallest mean square error compared with gradient pyramid algorithm, wavelet fusion algorithm and Contourlet fusion algorithm.