A Novel Deghosting Method for Exposure Fusion.

Chunmeng Wang,Chen He
DOI: https://doi.org/10.1007/s11042-018-6261-5
IF: 2.577
2018-01-01
Multimedia Tools and Applications
Abstract:A novel ghost-free exposure fusion method for generating an HDR image of a dynamic scene is presented in this paper. Given a sequence of input images with gradually increased exposures, due to the theory that the luminance is linearly depended on the exposure time (Mertens et al. Comput Graph Forum 28(1):161–171, 2009), each input image is normalized to make it have consistent luminance with a reference image. Then moving objects in the dynamic scene are detected using a modified difference method for further exposure fusion. Experiments and comparisons show that our method has advantage in deghosting when the reference image contains saturated regions and generate high-quality results with natural textures. Furthermore, our method has a largely improved timing performance compared with previous reference-guided methods.
What problem does this paper attempt to address?