Research On Acceleration Method Of Hdr Imaging Based On Guided Up-sampling

HDRI can provide more dynamic range and image details,which are widely used in consumer electronics,medical imaging,remote sensing,military reconnaissance and other fields.HDRI imaging needs complicated steps,such as reconstruction,tone mapping and image enhancement,so the complexity of the algorithm is high.In addition,the imaging efficiency and the application of HDRI are limited.Therefore,the research of HDRI imaging acceleration has gradually become a hot spot.The Super-Resolution algorithms based on single image can improve the efficiency of HDRI imaging,but it can not fully show the color and detail information at high resolution.Compared with the Super-Resolution algorithms based on single image,the Super-Resolution algorithms based on multiple images can obtain a high-quality image by up-sampling the Guidance Image,and the loss of image information is avoided.However,it has many limitations,and can only accelerate a specific step in HDRI imaging,which result of the improvement of overall imaging efficiency is limited,and the image processing time is relatively long.In this paper,a High Dynamic Range Imaging algorithm based on Joint Up-sampling is proposed,which can effectively reduce the number of pixels needed to be processed in High Dynamic Range Imaging,reduce the delay caused by the increase of processing steps in the imaging process,and achieve the overall acceleration of High Dynamic Range Imaging process on the premise of guaranteeing image quality.The specific work is as follows:(1)A High Dynamic Range Guidance Image is proposed.The Guidance Image can be extracted from several LDRIs to complete texture information of scene.It can be used for HDRI reconstruction based on Guided Up-sampling,and shorten the reconstruction time while ensuring the imaging quality.(2)Based on the Guidance Image,a High Dynamic Range Imaging algorithm based on Guided Up-sampling is proposed,which can effectively reduce the number of pixels input in High Dynamic Range Imaging and improve the imaging speed.The algorithm has no strict restrictions on the type and number of image operators.The algorithm has strong expansibility,which the image operators can increase or decrease according to the demand.Because of the noise during shooting,this paper also add the noise suppression ability to the proposed algorithm.The experimental results show that,on the premise of ensuring the imaging quality,the proposed algorithm can significantly shorten the time required for High Dynamic Range Imaging,and with the increase of resolution and imaging operators,the efficiency of the algorithm will be higher.In addition,image processing based on multi-exposure fusion technology can also be accelerated by the proposed algorithm,and good results can be obtained.