Underwater Image Enhancement Algorithm for Real Time Monitoring and Surveillance

Underwater Images are rarely being taken into consideration and capturing such images is big challenge. Due to attenuating and scattering properties of water, underwater images will be affected. Observing collected database leads to realization of high amount of Haze in an image which was the dominant degrading factor. Hence an enhancement and restoration of images captured is a need for accurate analysis. The proposed algorithm applies Single Image Dehazing to resolve the issue related to underwater image degradation. Firstly, calculation of dark channel to give low intensity value image is performed followed by finding the atmospheric light for underwater images. Further, the transmission image is obtained by using Gaussian Low Pass Filter, which reconstructs image in very less time by removing a good amount of haze as compared to Soft-Matting approach. Refined & Reconstructed image shows improvement in an image contrast according to the SNR values. Also the brightness values are well preserved than input image. Real time monitoring and surveillance in an underwater domain are the potential application areas of this method.