Research On Several Key Technologies Of Image Quality Improvement Based On Human Visual System

The promotion of image quality has been widely concerned.Because of the influence of camera,light source,lens,object material and other factors,the acquired images have different degrees of detail blur,color deviation and other problems,making the computer post-processing and identification more difficult,and even to the end user can cause difficult to solve the problem The Therefore,the quality of the image to enhance the important theoretical and practical significance.In this paper,we focus on the characteristics of color vision,specular reflection and so on,and combine the characteristics of human visual system to study the color correction and high light removal in image quality improvement.This is in the laboratory commitment to "based on human visual simulation of the online wall and floor tiles intelligent quality inspection and classification system","based on human visual reconstruction technology traffic video visibility detection method" and many other national,provincial projects more reflected.Based on the analysis of color correction and high light removal at home and abroad,this paper introduces the human eye physiological structure,human visual model,human visual characteristics,and proposes color correction and high light removal based on human visual characteristics.Firstly,the current color correction algorithm is analyzed,and the color correction based on the pattern search is proposed.Combined with the color logarithmic image processing to simulate the human eye color vision forming process,a color correction based on the visual characteristics of the human eye is proposed Finally,a detailed analysis of the current high light removal method,combined with the human eye visual brightness perception characteristics of the human eye based on the visual characteristics of high light removal.The experimental results show that the proposed method can achieve better results.The innovation of the paper lies in:· The color correction algorithm based on the pattern search optimization algorithm is proposed,and the CIEDE2000 chromatic aberration equation is best introduced.The traditional least squares analysis is improved and optimized by pattern search.The experimental results show that the algorithm is superior to the traditional polynomial regression· Color correction based on human visual characteristics is proposed.The pattern search algorithm finds the local optimal solution,and the initial value affects the correction result.Thus,inspired by the color perception of the human visual system,the image is corrected in the CLIP framework combined with visual multichannel and visual non-linearity.The experimental results are more consistent with human visual perception.· The high light removal algorithm based on multi-light source is improved,and the high light removal based on the human visual characteristics is proposed.The light and darkness adaptive function and the visual nonlinearity are combined to improve the light and dark adaptation function,so as to weaken the single image part of the high light,and finally use The median gradient reconstructs the image that has been weakened.The algorithm has high light removal effect and high robustness.