大雾天气对舰船在海上航行的安全造成极大干扰，同时也对雾天舰船的作战能力产生影响。大雾天气下获取的图像会存在舰船特征信息丢失、对比度失真等情况，本文提出基于视觉传达技术的舰船航行图像去雾增强方法，研究雾天的图像衰减模型，在此基础上研究直方图均衡法和多源目标融合图像去雾算法，并对比不同算法的图像去雾效果。提出基于视觉传达的船舶特征提取方法，实现视觉传达的雾天舰船航行图像监控系统，通过对图像的去雾及增强，改善了雾天下图像的视觉效果，提升了舰船航行的安全性。
Foggy weather greatly interferes with the safety of ships sailing on the sea, and also affects the combat ability of ships in foggy weather. The image acquired in foggy weather will have ship feature information loss, contrast distortion, etc. In this paper, an enhancement method for ship navigation image defogging based on visual communication technology is proposed, and the image attenuation model in foggy days is studied. On this basis, histogram equalization method and multi-source target fusion image intensity defogging algorithm are studied, and the image defogging effects of different algorithms are compared. A ship feature extraction method based on visual communication is proposed, and a ship navigation image monitoring system based on visual communication is realized. By de-fogging and enhancing images, the visual effect of images in foggy days is improved, and the safety of ship navigation is enhanced.
2024,46(7): 163-166 收稿日期：2023-6-25
DOI：10.3404/j.issn.1672-7649.2024.07.027
分类号：U667.65
作者简介：井新新(1990-),女,硕士,讲师,研究方向为视觉艺术设计及数字媒体艺术设计
参考文献：
[1] 张雪. 基于PSD和改进YOLOv5的雾天船舶检测方法研究[D]. 大连: 大连海事大学, 2023.
[2] 刘涛. 基于DCPDN和YOLOv4的大雾天气船舶目标检测研究[D]. 武汉: 华中科技大学, 2021.
[3] 张晓鹏, 许志远, 曲胜, 等. 基于改进YOLOv5深度学习的海上船舶识别算法[J]. 大连海洋大学学报, 2022, 37(5): 866-872.
ZHANG Xiao-peng, XU Zhi-yuan, QU Sheng, et al. offshore ship recognition algorithm based on improved YOLOv5 deep learning[J]. Journal of Dalian Ocean University, 2022, 37(5): 866-872.
[4] 宋佳怡, 谢维信, 王鑫. 融合暗通道滤波和空间金字塔的图像去雾算法[J]. 信号处理, 2019, 35(5): 816-824.
SONG Jia-yi, XIE Wei-xin, WANG Xin. Image dehazing algorithm based on dark channel filtering and spatial pyramid[J]. Signal Processing, 2019, 35(5): 816-824.
[5] 张成, 潘明阳, 高翊然, 等. 融合暗通道先验与循环生成对抗网络的航海图像去雾模型[J]. 大连海事大学学报, 2022, 48(4): 84-93.
ZHANG Cheng, PAN Ming-yang, GAO Yi-ran, et al. nautical image dehazing model based on dark channel prior and cyclic generation adversarial networks[J]. Journal of Dalian Maritime University, 2022, 48(4): 84-93.
[6] 张月. 基于改进YOLO算法的海面船舶目标智能检测研究[D]. 上海: 上海师范大学, 2020.
[7] 雷琴, 施朝健, 陈婷婷. 基于天空区域分割的单幅海面图像去雾方法[J]. 计算机工程, 2015, 41(5): 237-242.
LEI Qin, SHI Chao-jian, CHEN Ting-ting. Fog removal method of single sea surface image based on sky region segmentation [J]. Computer Engineering, 2015, 41(5): 237-242.
