您好,欢迎访问中国水产科学研究院 机构知识库!

Vision-based dual network using spatial-temporal geometric features for effective resolution of fish behavior recognition with fish overlap

文献类型: 外文期刊

作者: Zhao, Haixiang 1 ; Wu, Yuankai 3 ; Qu, Keming 2 ; Cui, Zhengguo 2 ; Zhu, Jianxin 2 ; Li, Hao 2 ; Cui, Hongwu 2 ;

作者机构: 1.Shanghai Ocean Univ, Coll Fisheries & Life Sci, Shanghai 201306, Peoples R China

2.Chinese Acad Fishery Sci, Yellow Sea Fisheries Res Inst, Key Lab Sustainable Dev Marine Fisheries, Minist Agr & Rural Affairs, Qingdao 266071, Shandong, Peoples R China

3.Tech Univ Munich, Sch Computat Informat & Technol, Munich Inst Robot & Machine Intelligence MIRMI, Dept Comp Engn,Chair Media Technol, D-80333 Munich, Germany

4.Laoshan Lab, Lab Marine Fisheries Sci & Food Prod Proc, Qingdao 266237, Peoples R China

5.Chinese Acad Fishery Sci, Yellow Sea Fisheries Res Inst, Qingdao 266071, Shandong, Peoples R China

关键词: Fish behavior recognition; Fish state recognition; Spatial-temporal graph convolutional networks; Slowfast networks; Micropterus salmoides

期刊名称:AQUACULTURAL ENGINEERING ( 影响因子:4.0; 五年影响因子:3.8 )

ISSN: 0144-8609

年卷期: 2024 年 105 卷

页码:

收录情况: SCI

摘要: In this study, a novel visualization framework for fish behavior recognition based on Slowfast networks and spatial -temporal graph convolutional networks (ST-GCN) is proposed. The framework can directly recognize fish behaviors in continuous videos and classify fish states in cases of severe fish stacking. A self -constructed fish behavior dataset containing 10 single fish HD videos and 300 fish schooling video clips covering three action categories and two state categories was collected. The evaluation was performed on this behavioral dataset. The results show that the framework achieves accuracies of 95.00% and 88.61% for state recognition and action recognition, respectively, exceeding those of several benchmark methods. Robustness and generalization experiments, as well as fish feeding experiments, were also conducted to demonstrate the potential application of the framework for guiding smart feeding in real production activities. The framework provides a novel solution for fish behavior analysis in the visual domain and can be extended to other aquatic animals or scenarios.

  • 相关文献
作者其他论文 更多>>