HDR Video Dynamic Mapping In Engineering Applied Research

With the rapid development of science and technology,people's demand for higher quality image film and television is becoming stronger.High dynamic range(HDR)images and films quickly enter people's daily life.HDR film and television adopt more advanced standard to bring people more real sensory experience.HDR images and films,in order to show natural scenes more truthfully,often exceed the dynamic range of common display devices at present.Therefore,it is better to map HDR images to ordinary display devices through algorithms.Based on the existing hue mapping algorithm,a detail-preserving hue mapping algorithm is proposed for HDR images,which can preserve more details of while mapping.Based on the proposed hue mapping algorithm,the simple application of the algorithm in HDR video is discussed.The following details the work:1.Based on the ATT algorithm of global hue mapping,an improved method is proposed.ATT algorithm is a global hue mapping algorithm based on histogram and human visual model proposed by Khan et al in recent years.In order to keep the details of the image better while dynamic mapping,edge preserving filtering is introduced,that is,fast bilateral filtering.By filtering the image stratification,the HDR gray image to be processed is divided into basic layer and detail layer.Basic layer contains the main dynamic information of the original HDR image,and the basic layer is dynamically compressed by human vision model.The original image and the basic layer are reduced to the detail layer image,the dynamic compressed basic layer is fused with the detail layer,and then the color can be restored to obtain the image that can be displayed in the commonly used display equipment,so as to facilitate people to observe and study.According to the subjective and objective evaluation,the proposed algorithm can map the dynamic information of the image well,and at the same time,the details of the bright and dark parts of the HDR image can be better preserved.There is no obvious transition compression problem,and to some extent avoids the image visual defects caused by artificial operation such as halo and gradient inversion,and has better image quality.2.After introducing this algorithm HDR video field,there are some problems such as time efficiency and time domain flicker.Aiming at the time efficiency of the algorithm,a histogram-based scene detection optimization algorithm is proposed.Usually a video scene consists of many similar image sequences.It can be considered to use a hue mapping for the same scene in the video to reduce the operation time.In addition,for the flicker problem caused by different frame mapping brightness of the same scene,a method of inter-frame fusion is proposed,and the picture distortion caused by inter-frame fusion of different scenes is avoided by scene detection.3.The global and local hue mapping algorithm and the objective evaluation method of the image are studied,and the techniques of edge preserving filtering,color space,video image processing and human visual model are also studied.After testing,the proposed algorithm can map the dynamic range of the HDR image well and keep the details of the image better.At the same time,the proposed video scene switching detection algorithm and interframe fusion algorithm also have good results,which can effectively solve the problems of time efficiency and time domain flicker.