Contrast Enhancement and Visibility Restoration of Underwater Optical Images Using Fusion

This paper proposes a novel method for underwater image enhancement algorithm using Principal Component Analysis (PCA) fusion. For fusion process, the dual images are derived using single underwater degraded image. The novelty of this work is the fact that it does not demand any underwater scene parameters beforehand and is independent of prior based information. Secondly, we demonstrated the successful use of homomorphic filter and adaptive histogram equalization for individual color channels followed with image smoothing as two inputs to fusion scheme. The fused image is subsequently post-processed using color constancy technique for effective results. The visual and quantitative analysis of this method is carried out with contemporary underwater dehazing and enhancement algorithms. Metrics such as entropy and quantification of restored edges is utilized to validate our findings. The outcome reveals better exposure of dark areas, enhancement of edges, increase in overall contrast and preservation of