基于大气散射模型的低照度视频增强算法

Low illumination video enhancement algorithm based on the atmospheric scattering model

低照度情形下光照条件复杂,摄像设备采集到的视频无法满足实际应用需求,因此,需要进行增强。提出一种基于大气散射模型的低照度视频增强算法。将视频分帧,对于每一帧图像,提取V分量。提出透射率和大气光值估计方法,更加准确地描述图像的照度分布,从而避免过曝光现象;根据低照度大气散射模型,求解得到初步增强的V分量。为了丰富图像细节,提出细节优化方法,得到增强图像,合成每一帧增强图像得到增强视频。对增强结果进行主观和客观分析,结果表明,所提出的算法在视觉效果和客观评价指标方面均优于对比算法,其中所提出算法的清晰度、边缘强度、基于块的对比度和结构相似度均排名第一,且算法运行速度较快。

Due to the complex conditions under low illumination, the video captured by camera device cannot meet actual application. Therefore, it needs to be enhanced. A low illumination video enhancement algorithm based on the atmospheric scattering model is proposed. Video is divided into frames, and for each frame image, the V component is extracted. The illumination distribution of the image is more accurately described by the proposed estimation methods of transmission and atmospheric optical value to avoid over-exposure. The initial enhanced V component is obtained by solving the low illumination atmospheric scattering model. To enrich the image details, a detail optimization method is proposed to obtain the enhanced image. Each frame of the enhanced image is synthesized to produce an enhanced video. A subjective and objective analysis of the enhancement results shows that the proposed algorithm is superior to the comparison algorithm in terms of visual effect and objective evaluation index. The proposed algorithm ranks first in terms of clarity, edge strength, patch-based contrast quality index, and structural similarity, and the algorithm runs faster.