Radiometric response functionsfrom differently illuminated images

In this paper, we propose a novelradiometric response function estimationalgorithm for high dynamic range imaging. In this algorithm, imagesarecaptured in random time interval of asame scene. The camera response function (CRF)is estimated based on each image and obtained the average CRF. Then, the ratio of eachCRF and the average value is definedas acontrol factor tofinally determine the optimal imageCRF. Wepresent an image fusion method to fuse all the images to generate an optimal high dynamic range image. Results on a variety of benchmarked and real images demonstrate that the proposed approach improves image quality and can obtain the best high dynamic range (HDR) image.