Study Of High Dynamic Imaging Algorithm Based On Multi-exposure Sequence

The dynamic range of natural scenes and that can be recognized by human visual system is far larger than the traditional imaging equipment.Due to the dynamic range of natural scenes is too large,traditional imaging equipment cannot directly capture all the details of the scene.High Dynamic Range(HDR)image has wide dynamic range,rich image detail and realistic visual effect,and the technologies for acquiring HDR image is very important in the field of digital image processing.The research of high dynamic imaging technology is composed of two major parts: HDR image synthesis and tone-mapping algorithm.The main HDR image synthesis method at present is multi-exposure image fusion,it takes different level of exposure images with the same scene,to retain as many image details as possible for different dynamic ranges in the same scene by the fusion algorithm,realizing the goal of HDR image.And the goal of tone mapping for HDR image is compressing the dynamic range of HDR image in order to display on ordinary displays and can be retain the image details information of the HDR images.This paper based on the deep study in the theory of the technologies for acquiring HDR image,one common drawback with the HDR image synthesis and tone-mapping algorithm is losing the image details,contrast and color information in the under-exposure and over-exposure region of the HDR image.To overcome this phenomenon,a novel image exposure region segmentation algorithm based on Local Binary Pattern and the feature of gray histogram is proposed in this paper.Then through the analysis of adaptive median filtering and connected domain Correcting the image is divided into over-exposure region,normal-exposure region and under-exposure region.In order to enhance the details in the over-exposure region and under-exposure region of image,the image exposure region segmentation algorithm is used for multi-exposure image fusion.And fusion of HDR image by calculates the weight matrix according to the exposure region local entropy,image saturation and image contrast.The image exposure region segmentation method is also integrated into tone mapping algorithm.To enhance the details of image,a piecewise linear mapping function is constructed to compress the dynamic range of different exposure region.Experimental results demonstrate that the method can obtain HDR image and tone mapping result image with better visual effect.It can enhance the saturation of image,while increases image details and local contrast in over-exposure and under-exposure region of image.