Feasibility study for using video detection system data to supplement automatic traffic recorder data.

PDFTech ReportFHWA-GA-15-111310406Video imaging detectorsTraffic surveillanceTraffic volumeTraffic dataRemote sensingData collectionAccuracyFeasibility analysisGeorgia. Department of Transportation. Office of ResearchGuin, AngshumanHunter, MichaelSuh, WonhoAnderson, JamesGeorgia Institute of Technology. School of Civil and Environmental EngineeringUnited States. Federal Highway AdministrationUS Transportation CollectionThe objective of this study was to determine the feasibility of incorporating Georgia NaviGAtortraffic volume data with Georgia Department of Transportation (GDOT) traffic volume data toenhance federal reporting. Some of the pertinent conclusions from this study are:1. Accuracy of video detection system (VDS) counts varied from site to site and lane to lane.2. Accuracy for VDS with gantry-mounted cameras was not significantly better than that ofpole-mounted VDS sites.3. Accuracy of VDS sites with cameras mounted on 36 ft. offset poles was marginally lowerthan units with cameras mounted on 24 ft. offset poles.4. Accuracy of counts from VDSs and remote traffic microwave sensors (RTMSs) is sensitive to site-specific deployment characteristics.5. There is a likely limit to the number of lanes that may be accurately counted by a singleVDS unit.6. Counts aggregated over all lanes provide the highest accuracy.7. Given the variability in data quality across detection stations, the use of VDS data to supplement GDOT data should be considered on a detector-by-detector basis (i.e. the datafrom a detector need to be individually checked for quality against ground truth data before they are used for federal reporting purposes)