Dual image fusion technique for underwater image contrast enhancement

Abstract

Underwater imaging is receiving attentions throughout these years. Attenuation of light causes the underwater images to have poor contrast and deteriorated color. Furthermore, these images usually appear foggy and hazy. In this paper, a new approach to enhance underwater images is proposed, which implements the integration of dehazing method, homomorphic filtering and image fusion. The dehazing method consists of multi-scale fusion technique, which applies weight maps in the pre-processing step. Homomorphic filtering and image fusion are then applied to the resultant image for contrast and color enhancement. Qualitative and quantitative evaluations are performed to analyze the performance of the proposed method. The results show the superiority of the proposed method in terms of contrast, image details, colors, and entropy. Moreover, implementation of Raspberry Pi with Picamera as standalone underwater image processing device is also successfully implemented.