基于反地理编码服务的内河船舶轨迹停留语义信息提取  

Associating Ship Position with Semantic Information by Using Reverse Geocoding Service

作　　者：黄亮[1,2] 闻宇 文元桥[1,2] 卢志刚 朱曼[1,2] 张治豪 HUANG Liang;WEN Yu;WEN Yuanqiao;LU Zhigang;ZHU Manu;ZHANG Zhihao(Intelligent Transportation Systems Research Center,Wuhan University of Technology,Wuhan 430063,China;National Engineering Research Center for Water Transport Safety,Wuhan University of Technology,Wuhan 430063,China;Zhejiang Scientific Research Institute of Transport,Hangzhou 311305,China)

机构地区：[1]武汉理工大学智能交通系统研究中心,武汉430063 [2]武汉理工大学国家水运安全工程技术研究中心,武汉430063 [3]浙江省交通运输科学研究院,杭州311305

出　　处：《中国航海》2022年第1期88-94,100,共8页Navigation of China

基　　金：国家自然科学基金(41801375;51679180);浙江省重点研发计划项目(2021C01010)。

摘　　要：由于内河水域电子航道图在完整性和及时性方面的不足,传统依赖电子航道图的船舶轨迹停留点挖掘方法在准确率和效率方面受到限制。在线地图资源蕴含着大量的地理空间语义信息,能为船舶活动轨迹提供丰富的地理关联解译。综合利用互联网中的地图数据服务,提出一种基于反地理编码的内河船舶轨迹停留语义信息提取方法。识别船舶在港口、码头等区域的轨迹停留段,提取每段的停留中心点;利用在线地图的反地理编码服务获取停留中心匹配的区域集合,计算每个区域的船舶停留特征;基于关键字对多个区域进行融合,生成船舶轨迹的停留语义信息。利用2018年7月—2018年8月长江区域船舶的真实轨迹数据进行试验,结果表明:该算法能有效地提取船舶轨迹的停留语义信息,平均准确率达到94%以上。The traditional electronic-chart relied way of mining data on ship’s port visiting is not reliable for inland waters, because the charts for those area is more likely incomplete or out of date. In contrary, abundant semantic information about geographic space is included in on-line maps which can supplement ship trajectory data with wide range of location-associated interpretation. A method to get the semantic information by using the reverse geocoding service is developed. The ship track segments where the ship does not move are identified. Those segments are picked up if they are within a port, or beside a pier and the center of each segment is extracted. The set of matching zones is obtained by using the reverse geocoding service provided by on-line maps. The semantic information associated with ship staying places is obtained through calculating characteristic of ship staying for each zone and conducting fusing according to the keywords. The method is verified against real data from the Yangtze River in the period of July and August, 2018 and achieved average accuracy was better than 94%.

关 键 词：船舶轨迹数据 反地理编码 船舶停留挖掘 船舶自动识别系统 

分 类 号：U675.79[交通运输工程—船舶及航道工程]