Underwater Image Restoration by Means of Blind Deconvolution Approach
Wei Zhang · Optics & Optoelectronic Technology · 2010
The degradation characteristic is one of the features of images acquired by range-gated underwater laser imaging systems for the reason of inevitable absorption and scattering by water. The underwater laser images have behaviors of low signal-to-noise ratio and blurry edge. In order to enhance the image quality, the degradation process of underwater laser imaging is analyzed. The basic noise pattern is studied, and with the partial knowledge of point spread function and modulation transfer function, the blind deconvolution approach is applied in underwater image restoration utilizing Wells' small angle approximation. The results of direct processing and using denoised image as the input of blind deconvolution are compared and discussed. Moreover, the different restoration results are obtained and discussed by intentionally changing the modulation transfer function and point spread function parameters. The restoration results show that the method proposed is effective in enhancement of underwater laser imaging to restrain background noise, clear object details and advance contrast.