首页|基于渔船轨迹数据的进出港区域识别方法

基于渔船轨迹数据的进出港区域识别方法

扫码查看
原文链接
  • 万方数据
  • 维普
针对当前渔船进出港区域获取方法成本高、更新周期长等问题，提出了一种基于渔船轨迹数据的渔船进出港区域识别方法。首先，提出基于多特征融合下轨迹点间相似性的轨迹划分算法，将渔船轨迹划分为不同渔船行为的轨迹段;然后，提出特征距离加权-K均值聚类算法(Feature Distance Weighted-K-means clustering algorithm，FDW-K-means)，将上一步得到的轨迹段特征作为聚类对象，实现渔船进出港轨迹段的提取。最后，综合运用DBSCAN(Density-Based Spatial Clustering of Applications with Noise)聚类算法和Del-Alpha-Shape算法对聚集的渔船进出港轨迹段轨迹点集进行边界提取获得渔船进出港区域。以椒江渔港和博贺渔港2021 年3 月的渔船轨迹数据为例，识别到椒江渔港和博贺渔港的渔船进出港区域的正确率分别为94。2%和95。8%。与使用K-means聚类算法或传统基于对各特征设定约束条件思想提取轨迹段的方法相比，该方法识别到的渔港渔船进出港区域正确率分别提高了10。7%，8。7%和9。5%，6。6%。实验结果表明所提方法能够有效识别渔船进出港区域，其结果能为渔船进出港监管提供科学参考。
Identification Method of Port Entering and Leaving Area Based on Fishing Vessel Trajectory Data
In view of the problems such as high acquisition cost and long update period,a new method based on fishing vessel trajectory data was proposed to identify port fishing vessel entering and exiting area.Firstly,a trajectory partitioning algorithm based on the similarity between trajectory points under multi-feature fusion was proposed.The fishing vessel trajectory data was divided into trajectory subsegments with different behaviors.Then,the Feature Distance Weighted-K-means clustering algorithm(FDW-K-means)was proposed to cluster the features of trajectory segment,and the trajectory segments of fishing vessel entering and leaving were obtained.Fi-nally,the DBSCAN clustering algorithm and Del-Alpha-Shape algorithm were used to extract the boundary of the gathered data set of the entering and leaving trajectory segment of fishing vessel to obtain the port fishing vessel entering and leaving area.Based on the fishing vessel trajectory data collected in March 2021 at Jiaojiang Fishing port and Bohe fishing port,the accuracy of fishing vessel entering and exiting area was94.2% and 95.8%,respectively.Compared with K-means or the traditional method based on the idea of setting the threshold value of features,the accuracy of extracting the entering and exiting area of fishing vessel in the two fishing ports was increased by10.7%,8.7% and 9.5%,6.6%,respectively.The experimental result shows that the proposed method can effectively extract the port entering and leaving area of fishing vessel,which can provide scientific reference for the supervision of fishing vessel entering and leaving the port.

fishing vessel trajectory datamulti-feature fusiontrajectory partitioningK-meansentering and leaving trajectory segmentport entering and leaving area

崔彤彤、徐硕、刘慧媛

展开 >

中国水产科学研究院 渔业工程研究所,北京 100141

渔船轨迹数据 多特征融合 轨迹划分 K-means 进出港轨迹段 进出港区域

国家农业科学数据中心渔业科学数据资源建设与共享服务项目中国水产科学研究院渔业工程研究所基本科研业务费专项渔业通信导航与大数据创新团队项目

NASDC2023XM032022HY-ZC0042020TD84

2024

10.20165/j.cnki.ISSN1673-629X.2024.0068

计算机技术与发展
陕西省计算机学会

计算机技术与发展

CSTPCD
影响因子：0.621
ISSN：1673-629X
年,卷(期)：2024.34(6)
  • 22