一种新的低照度彩色图像增强算法

摘要

为了增强彩色图像而不引起色彩失真,在HSV颜色空间中保持色相不变,提出了采用分段对数变换增强饱和度结合在多尺度Retinex算法的基础上,采用边缘保持增强色调的低照度彩色图像增强算法.实验结果表明,该方法在保持图像色相和图像边缘的情况下,显著改善了图像的视觉效果,提高了图像的亮度和对比度.25幅低照度图像的平均亮度、标准偏差和对比度分别提高了94.95％、20.93％和29.88％,相对于带色彩恢复的多尺度Retinex算法的熵和对比度增量分别提高了7.34％和151.51％,效果优于Retinex算法.%In order to avoid the color distortion during the image enhancement, this paper proposed a new low illumination color image enhancement method combining the saturation enhancement with segment logarithmic transformation and value en hancement with edge retaining( S_ Iog-V_edge) based on the multi-scale Retinex (MSR) algorithm without changing the hue in the HSV color space. Results show that this method can improve the low illumination images' visual effect and enhance the image brightness and contrast while retaining the image hue and edge. The 25 low illumination images average value, standard deviation and contrast incremental value are increased by 94.95% ,20. 93% and 29. 88% individually,the entropy and con trast incremental value are 7. 34% and 151.51% more than the multi-scale Retinex with color restoration ( MSRCR) .showing the effects are better than Retinex methods.