利用NSCT实现夜视图像的彩色化增强

摘要

针对夜视融合图像通常存在细节不够丰富、目标对比度低的问题,为了获得更为理想的图像增强效果,提出一种新颖的基于非下采样轮廓波变换(NSCT)的夜视图像彩色融合方法.构建了基于S函数与子图局部方差信息的可变加权融合策略,在NSCT域内实现可见光及红外源图像的自适应融合;将得到的融合图像与源图像进行组合并映射至YUV颜色空间,生成伪彩色融合图像;再运用颜色传递技术获得重染色的彩色融合图像.实验结果表明,该方法既能丰富彩色融合图像的细节,又能提高其亮度对比度和目标的可探测性,增强了观察者对场景的理解.%Traditional color night vision fusion methods always suffer from the problems of blurry visual effects and the low brightness contrast between the targets and the surroundings. To alleviate above problems, a novel fused image enhancement method for night vision is proposed in this paper. Firstly the nonsubsampled contourlet transform (NSCT) is performed on the infrared and the visible source images, respectively. Then a fused gray level image is obtained according to the self-adaptive fusion rules based on the S function and the local variance values. Subsequently, the source images and the fused image are synthesized together and mapped into the YUV color space, forming a pseudo-color image. Finally, a color transfer technique is employed to transform the pseudo-color image into a new version with natural color appearance. Experimental results demonstrate that the proposed method not only keep abundant details of the background, but also improve the target detectability and notably enhance the situation awareness.