The Research Of Low Illumination Video Enhancement Algorithm

We usually acquire images and videos in ideal luminance condition. But in some situations or for some special reasons, we’ll have to acquire images and videos in low luminance environment. The images and videos are terrible. It not only affects people’s watch, but also influences the effect of further processing. Thus, the low illumination video enhancement technology is very important.Firstly, we investigate the properties of image and video acquired in low luminance environment. They have three characteristics:first, the brightness of image is dark overall; second, compared with image acquired in ideal luminance condition, it has a large amount of noise; third, because of the limitation of luminance condition and light source itself, some components are lower than others among RGB components. Considered the properties of video acquired in practical low luminance environment and interframe correlation, a novel and complete low illumination video enhancement algorithm is proposed in this master’s thesis. It includes white balance, image denoising, image enhancement, edge compensation and interframe compensation. Through the practical test and comparing with other algorithms, we demonstrate the algorithm is effective. It can enhance low illumination videos effectively and simultaneously suppress the noise well. In addition, it can maintain edge of image and balance of color.