一种基于像素清晰度的NSCT域多聚焦图像融合方法

A method for multifocus image fusion in NSCT domain based on pixel clarity

针对非下采样Contourlet变换等多分辨率方法在图像融合过程中需要对源图像进行分解和重构,有可能改变源图像可靠信息等问题,提出了一种基于像素清晰度的NSCT的空域多聚焦图像融合方法.该方法根据HVS对局部对比度的敏感性和方向选择性,以及有用信息和噪声在NSCT域中的分布特点,引入局部方向对比度和方向向量的概念,用于衡量像素清晰度;然后通过比较多聚焦图像相应位置像素的清晰度,以选择模式决定融合图像的像素取自哪幅源图像.实验结果表明,该方法不仅能够从源图像中提取更多的有用信息,而且能够避免人为信息或高频噪声的引入,图像融合效果优于传统方法.

Considering the problem that NSCT-based muhiresolution fusion methods may change the reliable information during the process of decomposition and reconstruction, a muhifoeus image fusion method based on a new pixel clarity measure and NSCT in spatial domain was proposed. According to the characteristics of the human visual system （HVS） of frequency and direction selectivity, and the distribution difference between the image＇ s salient features and the noise in NSCT domain, the conception of local directional contrast and directional vector is introduced in the measurement of pixel clarity. Finally based on pixel clarity of the corresponding pixels of the source images, a selection principle is put forward for fusion. Experimental results demonstrated that the proposed method could not only extract more important visual information from source images, but also effectively avoid the introduction of artificial information. It significantly outperforms the traditional image fusion methods.