Improvement Of Multi-Scale Retinex Algorithm And Its Application In Video Image Enhancement

Nowadays,the images which are captured by Computer Vision Device contain a large number of low-illumination images,low-illumination images' gray value range is relatively narrow,leading to useful information in the image is difficult to distinguish,so people can't properly observe.When we want to identify the information we need from image,you have to enhance the low-illumination images.In this thesis,low-illumination images are selected as the research object,and the Retinex algorithm is used to enhance the images' quality and improve the visual effects.The thesis expounds the basic principle of Retinex and the current research status at home and abroad.By comparing and analyzing the common algorithms of Frankle-McCann Retinex(FMR),single scale Retinex(SSR)and multi-scale Retinex(MSR),it is found that the lack of the multi-scale Retinex algorithm is that the image does not take into account the grayscale nature of the pixel itself,resulting in over-enhancement of the high light part.There is a color distortion,affects the image visual effects.On the basis,the improved method is put forward in this thesis.The main steps of the improved method are as follows:firstly,the power function of the image is transformed to enhance the image contrast.Secondly,the filter is filtered by the bilateral filter to eliminate some of the noise in the image and protect the edge of the details of information.Then,using mean filter function to enhance the image instead of multi-scale Retinex Gaussian function.Finally,the genetic algorithm is used to select the function of power function and bilateral filter.In order to verify the effectiveness of the algorithm,some low-illumination images in different cases are chosen to experiment,and compare the processing effect with SSR and MSR.The image which uses improved algorithm reaches a more suitable brightness,retains most of the image details information and avoids the color distortion,so that the image screen is softer and color is more natural.After the image processing,we respectively select low-illumination video from the film and monitor,then to frame it and use the algorithm of this thesis to enhance,after that to restore the video.The method used in this thesis can effectively deal with low-illumination video images,make image quality and visual effects improved.