基于多传感器信息融合的智能交通信息语义描述

Semantic Description of Intelligent Traffic Information Based on Multi-sensor Information Fusion

针对摄像头采集交通信息时易受环境干扰且采集到的信息不够全面的问题,提出智能交通多传感器信息融合框架,对多种传感器采集到的交通信息进行融合。另外由于摄像头采集的视频数据量巨大且多为底层视觉信息,不便于用户信息检索,提出智能交通信息语义描述框架,对交通视频信息和传感器信息进行语义描述。实验结果表明,对多传感器采集到的信息进行融合能有效提高信息采集的精度,同时对交通视频信息进行语义描述将极大地方便用户对感兴趣信息的检索。

Considering the problems that when cameras are used to acquire traffic information their performance may be affected by environment and the information acquired is incomplete,a multi-sensor information fusion framework for intelligent transportation system（ITS） is proposed to fuse information gathered by different kinds of traffic sensors and cameras.On the other side,because the amount of information acquired by cameras is huge and most of the data is concerned with low-level visual information,a semantic description framework is proposed to describe the traffic information coming from traffic sensors and cameras for information retrieval of users.The results of experiments show that the fusion of the information collected by multiple sensors can improve the information acquisition accuracy and the semantic description for traffic video information can provide immense convenience for users to retrieve their interested information.