Research On The Algorithm Of High Dynamic Range Image Generation

HDR(High Dynamic Range,HDR)images have always been the focus of the display field since their appearance to the daily life of the public.Due to the physical limitations of the optical acquisition module in ordinary imaging equipment,the dynamic range that can be captured during real-time shooting is much lower than the real-time dynamic range,which makes it difficult to completely display the light and shadow contours of the real scene.The high dynamic range imaging technology based on multi-exposure can effectively overcome the limitations of imaging equipment hardware.This method uses a weighted fusion method to continuously shoot a group of images in real scenes,so that the images can record a higher dynamic range.This article conducts in-depth research on high dynamic range imaging algorithms and proposes two effective solutions.In this paper,an improved image blur enhancement algorithm based on Retinex theory is proposed.This method first uses the HSI color space to express the image;image brightness gain factor and brightness gain value is introduced based on Retinex theory.The purpose is to make the image brightness more meticulously enhanced,the color distribution is more uniform,and the scene in the picture is more realistic;The fuzzy enhancement operator is used to enhance the contour of the image scene.it can further deepen the contour of object in the image,preserve the details of the bright area of the image and improve the image contrast;Finally,the final image can be expressed in the RGB color space,so that the imaging results are easy to apply.Experimental results show that the algorithm can significantly enhance image details and textures,can completely restore dark areas of the image,and improve image quality.In this paper,a new multi-scale weight evaluation algorithm for color restoration and retinal enhancement imaging.First,the method uses the Retinex model with color restoration to estimate the brightness component of the original image,and combines the brightness compensation normalization function proposed in this paper to adjust the brightness of the original image,effectively increasing the brightness information of the main body of the image;Secondly,the method sets the four scales of exposure,saturation,contrast and color gamut to calculate the weighted value of the image to be fused,so as to achieve a more detailed evaluation of the image fusion weight;Finally,the Laplacian pyramid image reconstruction and fusion algorithm is used for the final processing,and the image sequence is weighting fused to obtain the imaging result.The experimental results show that the algorithm can greatly reduce the loss of image details,effectively improve the image expression effect,and has good superiority and application value in HDR imaging tasks.