Abstract
Various image enhancement methods have been proposed to make image better correlated to human visual perception. Generally, image enhancement can be performed in a local and/or global aspect. In this paper, we propose an innovative image enhancement framework consisting of bilateral tone adjustment (BiTA) and saliency-weighted contrast enhancement (SWCE). Different from most curve-based global contrast enhancement methods that enhance the bright and dark regions, BiTA furthermore enhances mid-tone regions that normally contain important scenes. Two implementation approaches of BiTA are presented: bilateral gamma adjustment (BiGA) and bilateral polynomial adjustment (BiPA). BiGA increases the overall “enhancement probability” and BiPA reduces the computational cost. For local contrast enhancement, SWCE integrates the concept of image saliency into a simple filter-based contrast enhancement method. Regions with higher saliency values, which indicates the regions have higher extent of human’s interest, deserve more degree of enhancement. In addition, saliency weighted relative entropy (SWRE) and saliency weighted relative entropy to noise (SWRE/N) are proposed to evaluate the enhancement quality. Simulation results show our proposed schemes achieve high contrast enhancement and great image quality.