无人船可用于海洋探测、军事侦查、军事打击等领域，结合GPS、AIS、无线通信技术以及图像视觉技术使得无人船实现智能化控制。无人船红外图像的单目视觉检测对实现目标检测以及跟踪具有至关重要的作用。本文提出使用YOLOV7算法对红外图像目标进行识别，确定红外图像训练集，对图像进行预处理后完成模型的训练及更新，最终实现对红外图像目标的识别。将KCF算法应用于红外图像目标跟踪，研究KCF算法对目标的跟踪流程，使用KCF算法和DCF算法进行仿真分析发现，同等情况下，KCF算法的跟踪准确率为78%，优于DCF算法。
Unmanned ships can be used in ocean exploration, military reconnaissance, military strikes and other fields, combined with GPS, AIS, wireless communication technology and image vision technology to achieve intelligent control of unmanned ships. The monocular vision detection of infrared images of unmanned ships plays an important role in target detection and tracking. This paper proposes to use YOLOV7 algorithm to recognize infrared image targets, determine the training set of infrared images, and complete the training and update of the model after image preprocessing, and finally realize the recognition of infrared image targets. KCF algorithm is applied to infrared image target tracking, and the tracking process of KCF algorithm is studied. Simulation analysis using KCF algorithm and DCF algorithm shows that the tracking accuracy of KCF algorithm is 78%, which is better than DCF algorithm in the same case.
2024,46(7): 159-162 收稿日期：2023-11-6
DOI：10.3404/j.issn.1672-7649.2024.07.026
分类号：U667.65
基金项目：湖北省教育厅科学研究计划指导性项目(B2018417)
作者简介：熊守丽(1985-),女,硕士,副教授,研究方向为大数据及计算机视觉
参考文献：
[1] 金久才, 张家林, 刘德庆, 等. 基于改进SiamRPN的高速无人船海上目标视觉跟踪研究[J]. 海洋科学进展, 1-10.
JIN Jiu-cai, ZHANG Jia-lin, LIU De-qing, et al. Research on visual tracking of marine targets by high-speed unmanned ship based on improved SiamRPN[J]. Progress in Marine Science, 1-10.
[2] 李可染, 李立刚, 贺则昊, 等. 基于岸堤特征提取的近岸无人船激光SLAM方法[J]. 激光与光电子学进展, 1-20.
LI Ke-ran, LI Li-gang, HE Ze-hao, et al. Laser SLAM method for nearshore unmanned ship based on embankment feature extraction[J]. Progress in Laser and Optoelectronics, 1-20.
[3] 葛燕龙, 金久才, 李立刚, 等. 视觉辅助下的无人船双雷达障碍物融合检测技术[J]. 舰船科学技术, 2023, 45(20): 87-92.
GE Yan-long, JIN Jiu-cai, LI Li-gang, et al. Dual radar obstacle fusion detection technology for unmanned ship with visual aid[J]. Ship Science and Technology, 2023, 45(20): 87-92.
[4] 谭静仪, 潘进, 黄婷, 等. 多变光照下的无人船图像模糊纹理光点定位识别[J]. 舰船科学技术, 2021, 43(12): 61-63.
TAN Jing-yi, PAN Jin, HUANG Ting, et al. Positioning and recognition of blurred texture light spot in unmanned ship image under variable illumination[J]. Ship Science and Technology, 2021, 43(12): 61-63.
[5] 刘雨青, 姜亚锋, 邢博闻, 等. 基于无人船装置的大水域环境监测系统设计[J]. 船舶工程, 2019, 41(1): 17-22+73.
LIU Yu-qing, JIANG Ya-feng, XING Bo-wen, et al. Design of large water area environmental monitoring system based on unmanned ship device[J]. Ship Engineering, 2019, 41(1): 17-22+73.
[6] 王贵槐, 谢朔, 初秀民, 等. 基于深度学习的水面无人船前方船只图像识别方法[J]. 船舶工程, 2018, 40(4): 19-22+99.
WANG Gui-huai, XIE Shuo, CHU Xiu-min, et al. Image recognition method for front ship of unmanned surface ship based on deep learning[J]. Ship Engineering, 2018, 40(4): 19-22+99.
