Efficient enhancement of underwater images using color correction and fusion

Abstract

This paper proposes an enhancement technique for underwater images that are degraded due to scattering and absorption. The method comprises of channel compensation and white balancing steps, and the output is given in parallel to the gamma correction and sharpening to adjust the contrast and remove high frequency irregularities. Using suitable fusion technique, the images are combined together, after which the quality metrics are evaluated. Measures are adopted to reduce the computational complexity of the method and to automate the execution. This gives user the flexibility to test the images and collect results with almost 50% less computation time. The method does not require any prior knowledge/training or channel model. Such features can be very crucial in time constrained or power limited applications.