基于NSST域人眼视觉特性的图像融合方法

Technique for image fusion based on NSST domain and human visual characteristics

针对多传感器图像融合问题,提出一种基于非下采样Shearlet变换域与人眼视觉特性的图像融合方法.采用非下采样Shearlet变换对源图像进行多尺度、多方向稀疏分解,得到低频子带图像和一系列不同尺度、不同方向的高频子带图像;提出一种视觉敏感度系数作为各子带图像融合的考量标准完成对源图像各对应子带图像的融合,设计了基于非下采样Shearlet变换与人眼视觉特性的图像融合算法,并采用非下采样Shearlet逆变换获得最终融合图像.仿真结果表明:该方法不仅拥有更理想的融合效果,还具有较高的运行效率.

A novel technique for image fusion based on the non？subsampled shearlet transform （NSST） domain and human visual characteristic （HVC） is proposed to resolve the problem of the multi？sensor image fusion. Multi？scale and multi？directional sparse decompositions of source images are performed by NSST, so that the low？frequency sub？images and a series of high？frequency ones with diverse scales and directions can be obtained. Then, as the e？valuation norm of sub？images fusion, the definition of visual sensitivity coefficient is presented to complete the fu？sion process of sub？images from each corresponding source image, respectively. Meanwhile, the algorithm for image fusion based on NSST and HVC is devised. The final fused image is achieved by utilizing inverse NSST to all fused sub？images. Experimental results show that the technique proposed has better performance, and higher running effi？ciency.