The images taken underwater are affected by the scattering of suspended particles,and may have adverse conditions such as blur,low contrast and color distortion,which affect the image quality.In order to enhance the visual effect of underwater im-ages with artificial intelligence technology,the research process proposes the dual residual mixed attention mechanism,and estab-lishes the corresponding image enhancement algorithm model on this basis.The use of the open source data set was used to test the performance of the algorithm model,and four control algorithms were set for it.The results showed that the algorithm performed the best in the two evaluation indexes of structural similarity and peak signal-to-noise ratio,and achieved the expected goal.
关键词
双重残差混合注意力网络模型/水下图像增强算法/性能试验
Key words
dual residual hybrid attention network model/underwater image enhancement algorithm/performance test
万方数据