An Underwater Image Enhancement via Wavelet domain Gradient Guided Filter

Pictures confined in underneath are often yield limited visibility and low dissimilarity due to haze in undersea. Existing approaches enhance pictures but frequently undergo noise issue; this paper presents a hybrid method for solving mage enhancing difficulty in frequency domain. Firstly, we propose locally adaptive Non locally robust regularization to deblur the image. The deblurred image has small gray-level rate in any color channel. Secondly we used an open dim channel scheme to increase visibility in low-intensity rate. Thirdly, gradient guided filter to enhance the details. Later, we use the soft-thresholding process to decrease noise in high-intensity rate to advance texture information. Finally, image is well enhanced via wavelet domain gradient guided filter. The projected technique intends to raise perceptual visibility, keep extra texture information as well lower noise effect. The performance evaluations prove that projected scheme give up better results by existing methods.