Underwater image quality assessment based on human visual system

With the increasing demand of underwater vision, more and more attention is paid to underwater image processing and it becomes very significant to improve the image perception effect Aiming at degradation and color bias of underwater images, a method for underwater image quality assessment (IQA) is proposed. The paper selected three indexes: sharpness, contrast and chroma, which are related to human vision system. Sharpness is calculated from the gradient values of pixels in the transverse, vertical and diagonal directions of the image. Contrast is calculated by the relative standard deviation of RGB three channels. Chroma is calculated through the relative value of red channel to green channel and blue channel. The weighted sum of the three indexes constituted the underwater image quality evaluation method in this paper. The proposed algorithm is compared with BRISQUE, NIQE, UIQM and UCIQE. Its performance is obtained. The experimental results show that the proposed underwater IQA method is better consistent with human visual perception, and can better reflect the quality of underwater image.