Low Intricacy‎ Multistage Algorithm for Underwater Image Enhancement

Abstract

Humanity currently lives in a technological era that witnesses rapid progress in multiple fields. Digital image processing is one of the modern technologies that has provided practical answers to many challenges including image enhancement, analysis, reconstruction, recovery, compression, processing, and understanding. One of these notable challenges relates to underwater photography. Underwater images are always exposed to less-than-ideal conditions due to environmental and physical factors. These include refraction of light in water, scattering of particles and dust in the aquatic medium, lack of illumination in deep water, and poor contrast. These challenges make it extremely difficult to analyze and extract valuable information without advanced processing. In this study, an improved color balance-fusion algorithm is provided by improving the image visuality and modifying some equations to obtain sharper and clearer images. The proposed algorithm begins by finding the white balance of the input RGB color image, after that, it improves the intensity. Next, the edges are improved using Gamma separately. The weights are then found for each image and combined to find naive fusion. The resulting image is processed using a color retrieval algorithm to produce the final image. along with comparisons to eleven other algorithms with various processing methods. Experimental results showed that this algorithm can significantly improve underwater images, increasing image clarity and making colors clearer. The improvement rates reached 5.8389 and 2.6778 for UISM and UICM metrics, respectively.

Keywords

• underwater
• concepts
• image enhancement image processing
• color images