Image Processing For LDR To HDR Image Conversion Using Deep Learning
D V Soundari, Rabin G, S Rishiyugan, Pravinkumar P · 2023
High Dynamic Range images and High Dynamic Range displays are getting popular nowadays. To attain the benefit of the High Dynamic Range, a lot of Low Dynamic Range images are to be converted to images with better ranges. Due to the reflection, the image may lose its visibility in some specific area and such area is called a highlight. The information will not be clear for the areas in highlight. The highlight should be removed in order to increase the image's visibility. It is not easy to obtain an image having dynamic range at high level for most of the people as it requires a pricey tool which can be used to function beyond the parameters that are present in the sensor technology of the conventional camera. Many low dynamic range images are combined to create a single HDR image. In the existing system the bright regions are brightened and the dark regions are darkened and the ranges with the extended brightness are merged in order to construct an image with High Dynamic Range. For Example, while capturing a crime scene in the camera with certain details that are absent in the image captured, the criminal investigation could be handled better by reconstructing the lost details of the image using this strategy. The exposure of the image is increased along with the brightness of the pixels that are highly dense and the output image is constructed with the properties necessary for a HDR image.