Cuckoo Optimization Algorithm based Image Enhancement

This paper proposes an extension to approach proposed in [13] for image enhancement using a combination of fuzzy logic technique and bio-inspired optimization algorithm. The transformation of the image data from RGB to HSV space has been done without altering HUE information. The image has been categorized into three regions with well tuned membership functions: underexposed, overexposed and mixed region on the basis of two threshold values. Gaussian membership function finds good suitability for fuzzification of overexposed and underexposed regions and mixed region is kept untouched which are further modified by a parametric sigmoid function. To get the quantitative analysis of the image; quality measures like fuzzy contrast, contrast and visual factors have been utilized. An objective function involves entropy and visual factor which is being optimized by bio-inspired optimization algorithm. Here, Cuckoo Optimization Algorithm (COA) has been used for parameter optimization and its results have been compared with the ACO based image enhancement on the scale of visual factor and execution time. COA based image enhancement found better than other approaches. The time taken to enhance the image has also been reduced as compared with latest approaches.