基于奇异值分解和引导滤波的低照度图像增强算法

Low Light Image Enhancement Algorithm Based on Singular Value Decomposition and Guided Filtering

针对现有低照度图像增强算法在处理图像后容易出现色彩失真、细节丢失、过度增强等问题,提出一种基于奇异值分解和引导滤波的低照度图像增强算法。首先通过Max-RGB模型获得初始光照分量,使用奇异值分解和引导滤波对初始光照分量进行优化,得到最终光照分量。利用Retinex模型,将原低照度图与光照分量图逐点相除,得到增强图像,并使用原始图像的绿色分量图作为引导图像,使用引导滤波对增强图像进行去噪处理。实验结果表明,提出的算法能够得到色彩更加真实、视觉效果更好的图像,同时能够避免过度增强、出现光晕等问题。

The existing low light image enhancement algorithms was prone to color distortion,detail loss and excessive enhancement after image processing.A low light image enhancement algorithm based on singular value decomposition and guided filtering was proposed.First,the initial illumination component was obtained by the Max-RGB model,and the initial illumination component was optimized by singular value decomposition and guided filtering to obtain the final illumination component.By using the Retinex model,the original low illumination image and the light component image was divided point by point to obtain an enhanced image.The G component map of the original image was used as the guide image,and the enhanced image was denoised by the guide filtering.The experimental results show that the algorithm can obtain more realistic images with better visual effects,and avoid problems such as excessive enhancement and halo.