一种新的彩色图像幸运区域融合算法

A Novel Lucky-Region Fusion Algorithm for Color Images

幸运区域融合是一种可以有效地恢复高分辨率的光学图像的方法。但已有的幸运区域融合算法通常依赖于固定分块选图,或者是使用多帧短曝光图像求偏微分方程以得到最终图像。这些方法需要将图像固定分块或是需要较大的计算量,无法满足实际应用场景的实时处理要求或是无法准确找到形状不规则的幸运区域。提出了一种可以不依赖于固定分块且不需要求解偏微分方程来获得融合图像的算法,即利用边缘指示函数和增益函数寻找幸运区域后进行直接融合处理。使用上述算法对实际拍摄的受到湍流影响的文字目标灰度图像以及具有景深的彩色图像进行了图像融合实验,结果表明,经算法处理得到的融合图像质量以及得到图像融合处理的速度有明显的提升,且可以实现在彩色图像上无色差地提高因大气湍流而退化的图像的分辨率。

Lucky region fusion is an effective method to restore high resolution optical images.However,the existing lucky region fusion algorithms usually rely on fixed block image selection,or use multi frame short exposure images to solve the partial differential equation to get the final image.These methods need to fix the image into blocks or require a large amount of computation,which cannot meet the real-time processing requirements of practical application scenarios or accurately find the lucky areas with irregular shapes.This article proposes an algorithm that does not rely on fixed partitioning and does not require solving partial differential equations to obtain fused images,which uses edge indication functions and gain functions to find lucky regions and directly fuse them.The results show that the quality of the fused images and the speed of image fusion processing obtained by the algorithm in this paper have been significantly improved,and the resolution of images degraded by atmospheric turbulence can be improved on color images without chromatic aberration.