Dual Domain Approach for Colour Enhancement of Underwater Images

Abstract

Scattering and Absorption of light in water are the two main problems in underwater scenario which result in low contrast, high level of blurring and dull colours in underwater images. To deal with these issues and to provide better visual quality image, the present paper proposes contrast stretching in approximation band of discrete wavelet transformed image. This is followed by intensity adjustment of different colour channels in spatial domain. To further improve the colour quality, image is processed in HSV (Hue, Saturation and Value) colour space. Different level of histogram stretching is applied on the R, G and B colour channels depending on their composition in underwater image. Result analysis indicates better results as compared to the state of art methods. Subjective results of proposed method show minimization of the bluish-green effect and reduction of non-uniform illumination leading to enhanced colours and image details. Quantitative results show that UIQM (Underwater Image Quality Measure) values between 1and 2 are achieved which signifies that images have been enhanced efficiently. Also, the Entropy values between 7 and 8 depict improvement in image details. Present method has application in underwater scenario for detection of corroded and leaked subsea pipelines.