DMCA
WB3.1 A Multimodal Framework for Vehicle and Traffic Flow Analysis
BibTeX
@MISC{Ploetner_wb3.1a,
author = {Jeffrey Ploetner and Mohan M. Trivedi},
title = {WB3.1 A Multimodal Framework for Vehicle and Traffic Flow Analysis},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Abstract—This paper presents an overview of a novel multimodal system being developed at UC San Diego for vehicle detection and traffic flow analysis. A Distributed Multimodal Array (DiMMA) framework is presented for sensory data acquisition, processing, analysis, fusion, and “active ” control mechanisms needed to recognize objects, events, and activities which have multi-modal signatures. Current sensing modalities being researched include video, audio, seismic, magnetic, and passive infrared. Feature extraction and data fusion techniques are being investigated to improve robustness and study the advantages and disadvantages of each sensing modality. Preliminary results of this rapidly deployable system are discussed, along with possible future expansions, including laser range scanners, geophones, pneumatic road tubes, and traditional inductive loops. T I.
Keyphrases
traffic flow analysis multimodal framework laser range scanner distributed multimodal array preliminary result feature extraction possible future expansion active control mechanism multi-modal signature data fusion technique sensory data acquisition pneumatic road tube traditional inductive loop vehicle detection uc san diego deployable system novel multimodal system
