基于本体的道路标牌自动审核

发布时间：2018-02-12 06:19

本文关键词： 语义技术路标本体合法性检测　出处：《武汉大学学报(理学版)》2015年04期 　论文类型：期刊论文

【摘要】：随着城市规模的不断扩大,道路标牌的有效管理是智能交通中的一个关键问题.本文在基于本体有机整合海量、多源、异构语义数据的基础上,研究基于国家路标设置规范的路标牌位置和内容有效性自动审核,是对语义技术在智能交通应用中的有益探索.首先,对地缘信息中的道路、兴趣点及路标建立统一模型,采用本体技术把海量数据在语义层面上关联起来.其次,对国家路标设置规范中的规则进行分类,对每一类规则分别构造规则模板及为其构造带参数的SPARQL查询,从而将路标设置指南进行规范形式化表示,在此基础上实现对路标牌的自动检测.最后,在构建的实验数据集上进行了实验,实验结果证明,提出的方法路牌检测准正确率达到85.1%,召回率达到92.1%.
[Abstract]:With the continuous expansion of urban scale, the effective management of road signs is a key problem in intelligent transportation. This paper integrates massive, multi-source and heterogeneous semantic data based on ontology. It is a useful exploration for the application of semantic technology in intelligent transportation to study the automatic verification of the location and content of road signs based on the norms of setting national signposts. Firstly, a unified model of roads, points of interest and landmarks in geo-information is established. The ontology technology is used to correlate massive data on semantic level. Secondly, the rules in the national road sign setting specification are classified, and the rules template and the SPARQL query with parameters are constructed for each rule, respectively. On the basis of this, the automatic detection of road signs is realized. Finally, experiments are carried out on the constructed experimental data set, and the experimental results prove that, The accuracy of the proposed method is 85.1 and the recall rate is 92.1.
【作者单位】：北京工业大学国际WIC研究院;江苏科技大学计算机科学与工程学院/智能交通研究所;荷兰阿姆斯特丹自由大学计算机系;日本前桥工业大学信息工程系;
【基金】：国家自然科学基金重点国际(地区)合作研究项目(61420106005)
【分类号】：TP391.1;U495

【参考文献】