| LCD has become a bridge between people and mobile devices,and is now the most widely used flat panel.At present,there are more and more applications of the mobile terminal display panel in the outdoors,and their size is bigger,the proportion of energy consumption is increasing.The demand for quality and energy consumption is also growing.The reduction of backlight(low-power mode)and the outdoor ambient illumination both will lead to low-quality display images.In recent years,image processing technology has been widely used in various fields,since image processing can effectively improve the image quality.LCD industry personnel are more and more concerned about the technology to achieve the purpose of display optimization.In this paper,the theoretical basis of LCD is introduced,and the influence of two factors of backlight and ambient light on display quality is analyzed.In the course of the study,it is found that most algorithms for backlighting or ambient light are not satisfactory in brightness,detail and color,and they cannot be well combined with the human visual system to reproduce the perception of the original input image.Tone mapping technology is to reconstruct the high dynamic range of the natural scene in the low dynamic range display device,so that the observer can obtain the same visual effect.Similarly,backlight and ambient light both lower the dynamic range of display devices.Based on the idea of tone mapping,the global and local operators are deeply studied in this paper,and the perceptual quality of the original image is reproduced with a lower dynamic range.The main work of this paper:(1)A fast global display optimized algorithm based on human perception is proposed.Most of the traditional algorithms directly deal with the brightness of the images,and cannot simulate the sensitivity of the human eyes well.This algorithm takes into account the adaptive response of human visual system,converting the brightness to the cone response value for processing.Then,the human visual perception model is used to keep the human visual perception differences.While the overall quality of the image is ensured,the human visual perception is retained.The algorithm has low complexity and good real-time performance,and is especially suitable for the environment with strong light.It is also easy to transplant on the hardware.(2)An optimized local tone mapping algorithm is put forward.The luminance information is mainly stored in the low frequency layer of the image,and the dynamic range compression occurs in the low frequency layer.Firstly,the image is divided into two layers by low pass filter,and an optimized function is designed in the low frequency,furthest conserves the global contrast and perceived brightness of the image.After obtaining a new low frequency layer,a new image is reconstructed from the high frequency layer which preserves the detail information of the image.Compared with the traditional image enhancement algorithms,the image appearance is better reproduced.The optimization function is solved by Lagrange theory,which reduces the complexity of the algorithm.Compared with the global algorithm,the local algorithm has higher computational complexity,but it is better to preserve the local details of the image,and can reduce the 30%-50% energy consumption while maintaining the image perception.(3)Considering different pixel values make different contribution on the image distortion,we designed a reasonable distortion definition in the local algorithm.This definition helps determine a new rational dynamic range for the base layer,solves the problem of pixel saturation in the traditional algorithm,which leads to the over enhancement and the loss of detail.(4)The subjective experiments and objective indexes are analyzed.The simulation results and experimental data show that the proposed algorithms can maintain the appearance of the image well,retain the perceptual contrast and perceived brightness,and reduce the energy consumption. |
PDF Full Text Request