基于学习机制的时空车辆跟踪与索引框架

Learning-Based Spatio-Temporal Vehicle Tracking and Indexing Framework

提出一个基于学习机制的时空车辆跟踪与索引框架,为ITS系统的车辆跟踪和索引提供了一个无人监管式解决方案。它通过引入了背景消隐方法、并行分区和类参数估计算法,获得对象间的时空关系,并使用多媒体扩大传输网络模型和多媒体输入字符串来表示它们的时空关系,并为其建模。

Proposes a learning-based spatio-temporal vehicle tracking and indexing framework. Introduces background subtraction method and SPCPE, it can obtain the Spatio-Temporal relation between objects. It also uses multimedia augmented transition network model, and multimedia input strings to index and model their spatio-temporal relations.