Performance improvement of enhancement methods for underwater images
Jasleen Kaur, Neelofar Sohi · 2017
Underwater images are highly important for studying various oceanic objects and underwater vegetation & life. Underwater scenes are usually veiled by the light interaction with the medium: low luminosity, absorption and many more. Capturing a clear and good quality underwater image is not an easy work, since underwater imaging usually suffers from unbalanced illumination, unbalanced pixel distribution, over-enhanced and under-enhanced areas, contrast and appearance of colors. Hence, study of underwater image enhancement techniques is necessary for oceanic application of computer vision. A large number of methods for enhancement of underwater images are available. In this study, few prominent enhancement methods are applied for enhancement of underwater images. Two methods are proposed based on BBHE, DSIHE, Gamma Filter, Butterworth Filter and color correction. Comparison of results produced by proposed methods is done with existing methods using performance metrics like RMSE, PSNR, Contrast, Entropy, Gradient. The proposed methods outperform the other methods.