Video Vehicle Detector Verification System (V2DVS)

This report describes a project which involved the development of an automated vehicle detector verification system designed to verify all individual detection results. The project was prompted by the need to assess the accuracy and attributes of the various types of roadway sensors and detectors currently in use. A traffic detector test bed was constructed on I-405 in Southern California by the California Department of Transportation. Because of the number of different types of detectors being tested, an automated testing process was designed. The design is based on the fusion of data from each of the detectors under test with data from a reference image processing system, in order to create a reliable composite ground truth record. The performance of the individual detectors was assessed by comparing it with the verified dataset. In addition, the report discusses the data acquisition architecture, data fusion methodology, computer vision detection methods, and automated data reduction and reporting methods.