Fusion Based Approach for Enhancement of Low - Illumination Imagery

Abstract

Photographs captured in dim-light conditions experience low visibility and low distinctness. Due to the poor visual aesthetics of photographs, the output of many algorithms’ degenerates. Due to their poor visibility, low-light images are not favorable for human examination. While several strategies for image enhancement have been suggested for low light photography, these approaches introduce under-enhancement or over-enhancement of contrast. In this paper a hybrid model has been proposed that focuses on low light photography exposure and contrast enhancement. Experiments demonstrated that compared with several different approaches, the proposed approach results in less contrast and less distortion. In turn, to improve low-light images, a comparison between them draws conclusions unique to creating a method. The proposed method works comfortably better than the existing system with respect to brisque and entropy as comparable metrics.