Research On Color Image Enhancement Algorithms In Low Light Conditions

There is no doubt that high-visibility images reflect the clear details of a target scene,which is critical for many vision-based technologies,such as object detection and tracking.However,images taken under low light conditions typically have lower visibility.Therefore,in order to improve the visual effect of images,it is of great theoretical significance and application value to carry out research on image enhancement algorithms.This thesis first introduces some basic theories of low illumination image enhancement,and then studies and analyzes some existing image enhancement algorithms,and improves and perfects the existing algorithms according to the defects and actual conditions of related algorithms.The specific research contents of this thesis are as follows:1.This thesis proposes an image enhancement algorithm based on content-adaptive histogram equalization.In order to prevent color from being affected during image enhancement,we choose to process an image in the HSI(Hue,Saturation,Intensity)color space.Firstly,the histogram components of intensity I are first redistributed.Secondly,since the global image enhancement cannot uniformly enhance the details of an image in different regions,in this case,an image needs to be locally enhanced,so an enhancement scheme of local contrast adjustment is adopted,and a Gaussian filter is used to eliminate the Chess board effect caused by local enhancement.Experimental results show that the enhanced algorithm proposed in this thesis effectively improves the contrast in the detail area of an image,and the enhanced image is more in line with the visual characteristics of the human eye.2.This thesis proposes an image enhancement algorithm based on dual-tree complex wavelet transform(DT-CWT)and Retinex theory.Because the HSV(Hue,Saturation,Value)color space is closer to the perception of color in the human visual system,it is convenient to process and recognize the color information.Therefore,we first chooses to process the low-illumination image in the HSV space,and use DT-CWT to process the V channel.Next,an improved local adaptive tone mapping method is applied to process the low frequency components of an image,and a soft threshold denoising algorithm is used to process the high frequency components of the same image.Then,we rebuild the V channel and adjust the contrast using the white balance method.Finally,the processed image is converted back to the RGB color space as an enhancement result.Experimental results show that the proposed method effectively improves the contrast and brightness of low-light images,which can achieve more natural visual effects than other algorithms.