Research on Underwater Image Denoising Based on Wavelet Threshold Method

Underwater images will produce severe noise during shooting, transmission and underwater backscattering, which seriously affects the accuracy of stereo matching and target positioning. For this reason, this paper proposes an improved wavelet denoising method. First, a new continuous threshold function is designed to overcome the problem that the soft threshold will weaken the image feature information and the hard threshold is not continuous at the threshold point. After that, considering the different propagation characteristics of signal and noise at different scales of wavelet transform, an adaptive threshold estimation method that changes with the decomposition scale is proposed. Experimental results show that the improved wavelet de-noising method can not only effectively filter the noise signal, but also retain the edge information of the image to the maximum extent, making the denoised image more conducive to subsequent stereo matching and target positioning.