交通场景中动态事件的语义表达方法  

A Method for Semantic Representation of Dynamic Events in Traffic Scenes

作　　者：于云[1] 曹凯[1,2] 

机构地区：[1]山东理工大学交通与车辆工程学院,山东淄博255049 [2]山东英才学院,山东济南250104

出　　处：《信息与控制》2015年第1期83-90,共8页Information and Control

基　　金：国家自然科学基金资助项目(61074140);山东省自然科学基金资助项目(ZR2010FM007)

摘　　要：针对动态交通场景中动态事件的语义表达方法不具有通用性的问题,提出了一个事件语义表达建模和分析的新方法.该方法将所有不同类型的信息汇集到分层概念模型中形成本体,利用描述逻辑语言描述交通场景中公理性和事实性的先验知识,并借用描述逻辑良好的推理能力来分析、表达动态场景中的事件,从而描述车辆的运动状态变化过程.应用案例的分析结果表明:分层本体的定义明确、拥有统一的框架结构,符合人类的认知过程,便于理解;描述逻辑语言的良好表达和推理能力增加了事件语义描述的可靠性.Semantic representation methods for dynamic events in traffic scenes are not universal. To solve this problem, we propose a new approach for event modeling and analysis with semantic representations. The ap- proach integrates all types of related information into a hierarchical conceptual model so as to form an ontolo- gy, and description logic is introduced to describe the axiomatic, factual knowledge in the scenes. Thus, the proposed approach uses a good reasoning ability to analyze and express the events in dynamic scenes; in this way the changes of vehicle movement states can be described. The analysis results of the application cases show that a layered ontology with clear definitions and a unified framework is in accordance with the human cognitive process, so it is easy to be understood by humans ; the good reasoning and expression ability of the description logic language strengthens the reliability of semantic expression.

关 键 词：动态场景 事件语义表达 分层本体 描述逻辑 

分 类 号：U121[交通运输工程] TP391[自动化与计算机技术—计算机应用技术]