基于水平相似度匹配机制的鱼群追踪与计数方法

doi:10.19678/j.issn.1000-3428.0069805

计算机工程 ›› 2026, Vol. 52 ›› Issue (3): 364-375. doi: 10.19678/j.issn.1000-3428.0069805

基于水平相似度匹配机制的鱼群追踪与计数方法

方斌¹, 邹青青²^,³, 巨浩飞¹, 鲍江辉²^,³, 段瑞⁴, 张东旭²^,³, 吕华飞²^,³, 王翔⁵, 许鹏飞¹, 段明²^,³^,*()

1. 西北大学信息科学与技术学院, 陕西西安 710127
2. 中国科学院水生生物研究所, 湖北武汉 430072
3. 中国科学院大学, 北京 101408
4. 武汉中科慧景科技有限公司, 湖北武汉 430075
5. 长江勘测规划设计研究有限责任公司, 湖北武汉 430019

收稿日期:2024-04-29 修回日期:2024-08-29 出版日期:2026-03-15 发布日期:2024-10-28
通讯作者: 段明
作者简介:
方斌, 男, 学士, 主研方向为深度学习、计算机视觉
邹青青, 学士
巨浩飞, 学士
鲍江辉, 博士
段瑞, 教授、博士
张东旭, 学士
吕华飞, 硕士研究生
王翔, 高级工程师
许鹏飞, 教授、博士
段明(通信作者), 研究员、博士
基金资助:
国家重点研发计划"智能传感器"重点专项(2022YFB3206900); 国家自然科学基金(32202914); 国家自然科学基金面上项目(32172955); 中国科学院"中央级科学事业单位改善科研条件专项资金"科研装备项目(GSZXKYZB2023019); 长江设计集团有限公司开放创新基金项目(CX2023K04)

Fish Tracking and Counting Method Based on Horizontal Similarity Matching Mechanism

FANG Bin¹, ZOU Qingqing²^,³, JU Haofei¹, BAO Jianghui²^,³, DUAN Rui⁴, ZHANG Dongxu²^,³, LÜ Huafei²^,³, WANG Xiang⁵, XU Pengfei¹, DUAN Ming²^,³^,*()

1. School of Information Science and Technology, Northwest University, Xi'an 710127, Shaanxi, China
2. Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan 430072, Hubei, China
3. University of Chinese Academy of Sciences, Beijing 101408, China
4. Wuhan Zhongke Huijing Technology Co., Ltd., Wuhan 430075, Hubei, China
5. Yangtze River Survey, Planning, Design and Research Co., Ltd., Wuhan 430019, Hubei, China

Received:2024-04-29 Revised:2024-08-29 Online:2026-03-15 Published:2024-10-28
Contact: DUAN Ming

摘要/Abstract

摘要：

鱼群多目标准确计数是水生态智能监测和集约化养殖产业中的重要环节, 对水域生态环境智能保护和水产养殖现代化具有重要作用。现有鱼群多目标准确追踪和计数方法主要适用于鱼群外观清晰、游速缓慢和方向稳定等较理想的情况, 难以有效适用于现实情况下存在的鱼群互相遮挡、游动迅速和方向多变等复杂情况。为此, 结合轻量化目标检测模型YOLOv5n, 提出基于水平相似度匹配机制的鱼群追踪与计数方法。将鱼群计数问题视为多目标检测与追踪问题, 设计水平相似度匹配机制, 并对SORT(Simple Online and Realtime Tracking)算法进行优化。通过高速水流中鱼群个体在帧与帧之间的位置关系对检测框中心点的水平距离进行限制, 以有效解决SORT算法存在的目标匹配混乱问题, 显著提高追踪效果。实验结果表明, 所提方法在鱼群多目标追踪数据集上的性能显著优于现有追踪方法, 对目标遮挡、方向变化等情况目标追踪性能提升显著, 并且该方法结构简单, 易于实际应用。

关键词: 多目标追踪, 鱼群计数, 水平相似度, SORT算法, 距离交并比

Abstract:

Multitarget accurate fish counting is crucial for the intelligent monitoring of water ecology and the intensive cultivation industry, playing a significant role in the protection of aquatic ecological environments and the modernization of fish farming. Existing methods for accurate tracking and counting fish with multiple targets are primarily suitable for ideal situations such as clear fish appearance, slow swimming speed, and stable direction. However, they often prove ineffective in complex real-life situations such as mutual occlusion, rapid swimming, and changeable direction of fish. Therefore, the lightweight target detection model YOLOv5n is combined and a method for tracking and counting fish based on the matching mechanism of horizontal similarity is proposed. This method regards the fish counting problem as a multitarget detection and tracking problem, proposes a horizontal similarity matching mechanism, and optimizes the Simple Online and Realtime Tracking (SORT) algorithm. The horizontal distance of the center point of the detection frame is limited using the position relationship between individual fish in the high-speed water flow to effectively solve the problem of target matching confusion in the SORT algorithm and significantly improve the tracking performance. The results show that the performance of the proposed method is significantly better than that of existing methods on a multitarget tracking dataset. Additionally, the target tracking performance significantly improves under the conditions of target occlusion and direction change. The proposed method has the advantages of simple structure and easy application.

Key words: multiple object tracking, fish counting, horizontal similarity, SORT algorithm, Distance Intersection over Union (DIoU)

方斌, 邹青青, 巨浩飞, 鲍江辉, 段瑞, 张东旭, 吕华飞, 王翔, 许鹏飞, 段明. 基于水平相似度匹配机制的鱼群追踪与计数方法[J]. 计算机工程, 2026, 52(3): 364-375.

FANG Bin, ZOU Qingqing, JU Haofei, BAO Jianghui, DUAN Rui, ZHANG Dongxu, LÜ Huafei, WANG Xiang, XU Pengfei, DUAN Ming. Fish Tracking and Counting Method Based on Horizontal Similarity Matching Mechanism[J]. Computer Engineering, 2026, 52(3): 364-375.

https://www.ecice06.com/CN/Y2026/V52/I3/364

图/表 14

图1 用于鱼群计数的过鱼通道实景图

Fig.1 Realistic scenery of fish passage for counting fish

图2 基于传统SORT算法的追踪效果图

Fig.2 Tracking rendering based on traditional SORT algorithm

图3 基于水平相似度匹配机制的鱼群追踪模型框架

Fig.3 Framework structure of fish tracking model based on horizontal similarity mechanism

图4 DIoU_Match原理示意图

Fig.4 Schematic diagram of DIoU_Match

图5 DIoU_Match追踪效果图

Fig.5 DIoU_Match tracking rendering

图6 Horizontal_Match追踪效果对比图

Fig.6 Comparison of Horizontal_Match tracking rendering

图7 Horizontal_Match原理示意图

Fig.7 Schematic diagram of Horizontal_Match principle

图8 追踪器的FPS和MOTA指标

Fig.8 MOTP and MOTA indicators of trackers

图9 DeepSORT中ID为30的鱼苗追踪效果图

Fig.9 Schematic diagram of ID=30 fish fry tracking rendering in DeepSORT

图10 SORT_DIoU追踪出现的问题

Fig.10 The problems encountered in SORT_DIoU tracking

图11 HS-SORT对横向距离做出的限制及其追踪效果图

Fig.11 HS-SORT′s limitations on lateral distance and its tracking rendering

图12 不同算法鱼苗数量小时连续帧追踪效果图

Fig.12 Continuous frame tracking rendering of different algorithms when the number of fish fry is small

图13 不同算法鱼苗数量大时连续帧追踪情况

Fig.13 Continuous frame tracking when the number of fish fry in different algorithms is large

参考文献 24

1	周典卓, 谭鹤群, 朱明, 等. 采用边缘曲率进行图像分割的高密度鱼苗计数方法. 农业工程学报, 2023, 39 (21): 258- 268.
	ZHOU D Z , TAN H Q , ZHU M , et al. Method for high density fry counting by image segmentation based on edge curvature. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2023, 39 (21): 258- 268.
2	时增林, 叶阳东, 吴云鹏, 等. 基于序的空间金字塔池化网络的人群计数方法. 自动化学报, 2016, 42 (6): 866- 874.
	SHI Z L , YE Y D , WU Y P , et al. Crowd counting using rank-based spatial pyramid pooling network. Acta Automatica Sinica, 2016, 42 (6): 866- 874.
3	BEWLEY A, GE Z, OTT L, et al. Simple online and realtime tracking[C]//Proceedings of 2016 IEEE International Conference on Image Processing. Washington D. C., USA: IEEE Press, 2016: 3464-3468.
4	YU F W, LI W B, LI Q Q, et al. POI: multiple object tracking with high performance detection and appearance feature[C]//Proceedings of ECCV 2016. Berlin, Germany: Springer, 2016: 36-42.
5	LI D L , MIAO Z , PENG F , et al. Automatic counting methods in aquaculture: a review. Journal of the World Aquaculture Spciety, 2020, 52 (2): 269- 283.
6	WOJKE N, BEWLEY A, PAULUS D. Simple online and realtime tracking with a deep association metric[C]//Proceedings of 2017 IEEE International Conference on Image Processing. Washington D. C., USA: IEEE Press, 2017: 3645-3649.
7	LABUGUEN R T, VOLANTE E, CAUSO A, et al. Automated fish fry counting and schooling behavior analysis using computer vision[C]//Proceedings of IEEE International Colloquium on Signal Processing and Its Applications. Washington D. C., USA: IEEE Press, 2012: 255-260.
8	SHARMA S, SHAKYA A. Fish population estimation from underwater video sequences using blob counting and shape analysis[EB/OL]. [2024-01-02]. http://conference.ioe.edu.np/ioegc2014/papers/IOE-CONF-2014-61.pdf.
9	HERNANDEZ-ONTIVEROS J M , INZUNZA-GONZALEZ E , GARCIA-GUERRERO E E , et al. Development and implementation of a fish counter by using an embedded system. Computers and Electronics in Agriculture, 2018, 145, 53- 62. doi: 10.1016/j.compag.2017.12.023
10	RASMUSSEN C, ZHAO J, FERRARO D, et al. Deep census: AUV-based scallop population monitoring[C]//Proceedings of IEEE International Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2017: 2865-2873.
11	ZHOU Y Z , YU H , WU J F , et al. Fish density estimation with multi-scale context enhanced convolutional neural network. Journal of Communications and Information Networks, 2019, 4 (3): 80- 88. doi: 10.23919/JCIN.2019.8917888
12	GEIGER A , LAUER M , WOJEK C , et al. 3d traffic scene understanding from movable platforms. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 36 (5): 1012- 1025.
13	吴必朗, 柳春娜, 姜付仁. 基于DeepSORT算法的鱼道过鱼种类识别和计数研究. 水利水电技术, 2022, 53 (9): 152- 162.
	WU B L , LIU C N , JIANG F R . Species identification and counting of fishes in fishway based on DeepSORT algorithm. Water Resources and Hydropower Engineering, 2022, 53 (9): 152- 162.
14	AHARON N, ORFAIG R, BOBROVSKY B Z. BoT-SORT: robust associations multi-pedestrian tracking[EB/OL]. [2024-01-02]. https://arxiv.org/abs/2206.14651.
15	LIU S, LI X, LU H C, et al. Multi-object tracking meets moving UAV[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2022: 8876-8885.
16	ZHOU J L , JI D X , ZHAO J , et al. A kinematic analysis-based on-line fingerlings counting method using low-frame-rate camera. Computers and Electronics in Agriculture, 2022, 199, 107193. doi: 10.1016/j.compag.2022.107193
17	SHARIF M H, GALIP F, GULER A, et al. A simple approach to count and track underwater fishes from videos[C]//Proceedings of the 18th International Conference on Computer and Information Technology. Dhaka, Bangladesh: [s. n. ], 2015: 347-352.
18	ZHANG Y F, SUN P Z, JIANG Y, et al. ByteTrack: multi-object tracking by associating every detection box[C]//Proceedings of European Conference on Computer Vision. Berlin, Germany: Springer, 2022: 1-21.
19	万琴, 李智, 李伊康, 等. 基于改进YOLOX的移动机器人目标跟随方法. 自动化学报, 2023, 49 (7): 1558- 1572.
	WAN Q , LI Z , LI Y K , et al. Target following method of mobile robot based on improved YOLOX. Acta Automatica Sinica, 2023, 49 (7): 1558- 1572.
20	ZHENG Z H, WANG P, LIU W, et al. Distance-IoU loss: faster and better learning for bounding box regression[C]//Proceedings of the AAAI Conference on Artificial Intelligence. [S. l. ]: AAAI Press, 2020: 12993-13000.
21	ZHANG L , LI W S , LIU C H , et al. Automatic fish counting method using image density grading and local regression. Computers and Electronics in Agriculture, 2020, 179, 105844. doi: 10.1016/j.compag.2020.105844
22	SMITH K, GATICA-PEREZ D, ODOBEZ J, et al. Evaluating multi-object tracking[C]//Proceedings of 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2005: 36-36.
23	圣文顺, 余熊峰, 林佳燕, 等. 融合注意力与特征金字塔的小尺度目标检测算法. 计算机工程, 2024, 50 (1): 242- 250. doi: 10.19678/j.issn.1000-3428.0066724
	SHENG W S , YU X F , LIN J Y , et al. Small-scale object detection algorithm integrating attention and feature pyramids. Computer Engineering, 2024, 50 (1): 242- 250. doi: 10.19678/j.issn.1000-3428.0066724
24	LE J Y, XU L H. An automated fish counting algorithm in aquaculture based on image processing[C]//Proceedings of International Forum on Mechanical, Control and Automation. [S. l. ]: Atlantis Press, 2017: 358-366.

[1]	聂志勇, 阴宇薇, 汤佳欣, 涂志刚. 一种基于边界框关键点距离的框回归算法[J]. 计算机工程, 2023, 49(7): 65-75.
[2]	王毅然, 经小川, 贾福凯, 孙宇健, 佟轶. 基于多智能体协同强化学习的多目标追踪方法[J]. 计算机工程, 2020, 46(11): 90-96.

选择文件类型/文献管理软件名称

选择包含的内容

基于水平相似度匹配机制的鱼群追踪与计数方法

Fish Tracking and Counting Method Based on Horizontal Similarity Matching Mechanism

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 24

相关文章 2

编辑推荐

Metrics

本文评价

模态框（Modal）标题

选择文件类型/文献管理软件名称

选择包含的内容

基于水平相似度匹配机制的鱼群追踪与计数方法

Fish Tracking and Counting Method Based on Horizontal Similarity Matching Mechanism

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 24

相关文章 2

编辑推荐

Metrics

本文评价