High dynamic range imaging by a pseudo exposure fusion method based on artificial remapping

Abstract

High dynamic range imaging is a promising image enhancement technology, which can reconstruct a high dynamic range (HDR) image with a real visual perception from captured one (or several) raw low dynamic range (LDR) image. However, the input LDR image only provides partial information of the scene. Besides, in traditional methods that require multiple captured images as input, it is prone to halos and/or ghosting artifacts due to camera shake or object motion. In this paper, we propose a simple yet valid method that firstly uses artificial remapping to generate a set of simulated multi-exposure image sequence from a single LDR image, then fuses them into a high-quality HDR image via multiscale pyramid framework-based exposure fusion. It was demonstrated that artifacts-free and details-preserving HDR image can be achieved. The method thus can be potentially applied to the field of high dynamic range imaging, and can be further used for astronomical exploration, biomedical imaging, and optical imaging fields in the future.