Research On Enhancement Algorithms Of Low-Light Image And Video Based On Retinex Theory

Image and video information plays a huge role in the current production and life.As the important information carriers,they promote the exchange of information and help people to understand the world more intuitively.However,in the night or dim lighting environment,the pictures and videos captured by the acquisition device often have a lot of noise,and some useful details will be directly lost.They will directly limit and affect the observation of the human eye,while it will seriously affect computer vision and image processing related applications.The purpose of this thesis is to effectively enhance low-light image and low-light video.On the basis of analyzing the traditional enhancement algorithm,this thesis focuses on the theory of Retinex and logarithmic image processing(LIP)model,and proposes an effective low-light image enhancement.Additionally,based on the proposed low-light image enhancement algorithm,this thesis uses the principle of optical flow method to propose an effective low-light video enhancement algorithm.The main work and innovations of this thesis are as follows(1)Due to the influence of strong light,the images acquired at night have high contrast However,the over-enhancement is produced in bright region when contrast enhancement method is directly applied to such an image.Therefore,in order to solve the problem of over-enhancement of this part of the image with large contrast,a low-light image enhancement algorithm based on Retinex adaptive reflection component estimation and logarithmic image processing subtraction post-processing is proposed.The algorithm is mainly divided into two parts.The first part is the reflection component estimation,and the second part is the contrast enhancement based on the logarithmic image processing subtraction(LIPS)model.First,adaptive parameter bilateral filters are used to get more accurate illumination layer data,instead of Gaussian filter.Moreover,the weighting estimation method is used to calculate the adaptive parameter to adjust the removal of the illumination and obtain the reflectance by just-noticeable-distortion(JND)factor.Then,the LIPS method based on maximum standard deviation of the histogram is applied to enhance reflectance image adaptively.Experimental results demonstrate that the proposed method can effectively enhance the details of the bright areas of the image while effectively enhancing the darker areas of the image,and avoid the over-enhancement of the bright areas(2)For low light video,an effective video enhancement technology based on the low-light image enhancement method is proposed.Because the main target of the low-light image enhancement method in this thesis is high contrast images,in order to applied to low-light videos,this thesis modifies the adaptive factor into an adaptive parameter that is equally applicable to most low-light images,so that each frame of the video image has a better enhancement effect.If the improved low-light image enhancement method directly enhances each frame of the image,it will cause the brightness flick of the video.In order to solve this problem,the optical flow method is introduced to the method.Experimental results demonstrate that the proposed method not only has better enhancement effect on the enhancement result of single frame,but also the brightness flicker phenomenon of video is improved after the introduction of optical flow method.