Date of this Version
This thesis discusses the use of video detection as a means for train detection. It reviews works done to increase safety at and near Highway-Rail Grade Crossings (HRGCs), discusses existing and past forms of technology and methods used in train detection and arrival time prediction, and provides summaries of investigations of other detection technologies used for vehicle detection. It then goes into depth on the use of radar and video detection with Autoscope Machine Vision Processing for train detection. This thesis provides a methodology of data collection, data analyses, results, and conclusions of video detection for train detection.
Data analyses concluded that Autoscope video detection works for detecting trains and recovering acceptable data on their speeds. Data obtained by radar and Autoscope video detection can be used to reasonably predict train arrival times at HRGCs and alert motorists near HRGCs of upcoming train arrivals and departures. The conclusions reached with this research also identified future research needs that will assist in creating a robust system for detecting trains with video detection.
Advisor: Elizabeth G. Jones