一种基于Contourlet变换的多聚焦图像融合方法

摘要

A multi-focus image fusion method was proposed based on Contourlet transform.Firstly,the multi-focus image are decomposed to the domain of the Contourlet transform.Then,the image fusion is implemented in different subbands.The lowpass subband coefficients are fused by the improved fusion algorithm based on local energy.The highpass subband coefficient are fused by the integrated fusion arithmetic based on D-S evidence theory.The experimental results showed that the proposed fusion method can integrate the information and retain the features of sources image more effectively.%提出了一种基于Contourlet变换的多聚焦图像融合方法。首先采用Contourlet变换对不同聚焦的图像进行分解，然后在不同的子带中进行图像融合。低频子带采用基于区域能量改进的融合算法；高频子带采用基于D-S证据理论的融合算法。实验结果表明，该算法能够有效地融合源图像信息，保持源图像特征。