Image or Video Contrast Enhancement and Performace Analysis using Image Decomposition based Technique
Amitha Muralidharan, P R Geetharanjin, M J Josemartin · 2019
Due to poor lighting and atmospheric conditions, the images are often under or over-exposed which results in low contrast or quality. In this paper, an intrinsic decomposition algorithm is applied for image as well as video contrast enhancement. Intrinsic decomposition priors are used as constraints on decomposed illumination and reflectance layers to yield a better solution. It implies that neighbourhood pixels with same colour would have similar illumination and reflectance. Hence this method is designed as an optimization problem by diminishing an energy function. This decomposition model performs on HSV space in order to minimize computing complexity and avoid colour artifacts in enhanced images. Images with different lighting and atmospheric conditions are enhanced using this algorithm and quality is analysed using various image quality metrics. This method is also applied to video enhancement. Experimental results provide better enhancement results for both images and videos.