Underwater Image Quality Evaluation based on Human Visual System
Jinglu Guo · 2021
With increasing attention to underwater observation and utilization of marine resources, people's demand for underwater vision continues to increase. The processing of underwater images has received more and more attention. Underwater image quality evaluation plays a crucial role in underwater image processing technology and is also an important standard for underwater image enhancement. In this paper, we proposed a new no-reference underwater image quality evaluation method, which selects three indexes related to the human visual system: chroma, definition, and contrast. The chroma is calculated by saturation and the ratio of red channels to green and blue channels; the definition is calculated through the maximum gradient value and the detail content of the image; the contrast is calculated by the standard deviation of the image pixel value. The new underwater image evaluation index uses a linear combination of three attribute values. Experimental results show that this method can effectively evaluate the quality of underwater images based on the perception of human eyes.