Research On High Dynamic Image Generation Algorithm Based On Visual System

High Dynamic Rang Image(HDRI)has been widely used in various aspects of society with its High Dynamic and wide color field.Compared with the usual low dynamic image,it uses the floating-point data format to store the irradiance of the scene directly,which is the real reaction to the scene brightness.Whether in detail or color,HDRI is better.However,this kind of image is only applied in some high-end neighborhood,which is not used by ordinary consumers,so it is of great practical significance for the research.This paper mainly studies the technique of synthesizing HDRI based on multiexposure.Before multi exposure fusion,Retinex algorithm is first used for image enhancement preprocessing.However,the Retinex algorithm is not sensitive to the processing of high brightness regions.In this paper,the multi-scale Retinex algorithm is improved,especially the adaptive threshold processing whose purpose is to extract details from high brightness areas.Then the detailed information is fused into the output image to get the enhanced image.Finally,the enhanced images are fused by the sparse representation method,and the fusion rule based on the atomic usage rate is improved,to get the fusion rule based on the atomic average usage rate.The experimental results show that the algorithm is much better than the algorithm that is not improved in detail expression and color.