Unsupervised Contrast Correction for Underwater Image Quality Enhancement through Integrated-Intensity Stretched-Rayleigh Histograms

Abstract

The attenuation of light that travels through the water medium results the underwater image to suffer from several problems. Low contrast and color performance are the problems that resulting the image to loss important information. In addition, the objects in the image are hardly differentiated from the background. Consequences from these problems, this paper extend the methods of enhancing the quality of underwater image with the aim of improving the image contrast and increase the color performance. The proposed method consists of two stages. At first stage, contrast correction technique is applied to the image. The image is multiplied with a gain factor. The image histogram is divided into two regions at the mid-point and stretched towards the higher and lower intensity values. The composition of these two different intensities images produces contrast-enhanced image. At the second stage, the image is applied with color correction, where the image is converted into Hue-Saturation-Value (HSV) color model. Dividing and stretching of S and V components increase the image color. By considering the contrast and color performance of the output image, the proposed method outperforms the state-of-the-art methods