期刊
JOURNAL OF MARINE SCIENCE AND ENGINEERING
卷 10, 期 10, 页码 -出版社
MDPI
DOI: 10.3390/jmse10101513
关键词
underwater image enhancement; U-Net; nonlocal means denoising; image sharpening
资金
- National Natural Science Foundation of China [61501278]
An algorithm combining color correction and detail enhancement is proposed to address underwater image issues. By improving nonlocal means denoising algorithm and U-Net, the algorithm effectively enhances underwater images while retaining edge features and texture information. The proposed algorithm demonstrates remarkable results in enhancing underwater images.
To solve the problems of underwater image color deviation, low contrast, and blurred details, an algorithm based on color correction and detail enhancement is proposed. First, the improved nonlocal means denoising algorithm is used to denoise the underwater image. The combination of Gaussian weighted spatial distance and Gaussian weighted Euclidean distance is used as the index of nonlocal means denoising algorithm to measure the similarity of structural blocks. The improved algorithm can retain more edge features and texture information while maintaining noise reduction ability. Then, the improved U-Net is used for color correction. Introducing residual structure and attention mechanism into U-Net can effectively enhance feature extraction ability and prevent network degradation. Finally, a sharpening algorithm based on maximum a posteriori is proposed to enhance the image after color correction, which can increase the detailed information of the image without expanding the noise. The experimental results show that the proposed algorithm has a remarkable effect on underwater image enhancement.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据