Advanced Image Sensing Methods for Traffic Surveillance and Detection

This report describes the results of a PATH-funded study conducted by the Cal Poly Transportation Electronics Laboratory intended to assess advanced imaging technologies for potential application toroadway surveillance and detection. A major motivation for this study was the possibility of improvedvisibility under conditions of fog or dust using wavelengths longer than that of the visible spectral band. Technologies considered included ten types of infrared video cameras and one millimeter-wave still-frame imaging system. Evaluation criteria were structured to reflect the surveillance needs of traffic management center personnel and information needs of computer vision systems. The various imaging technologies are described in detail, and intrinsic merits and limitations of each are discussed. Field work is described in which video samples were acquired using each imaging technology. A pair of metrics of image information were developed and applied using computer image analysis to image sequences produced by each technology. Evaluation with respect to these metrics indicated that all of the advanced imaging technologies assessed generally provided less useful image information than conventional color CCD video cameras, except under conditions of very dense fog, in which some types of infrared imaging provided slightly more usable information. Overall conclusions were that advanced imaging methods appear to be justified only for selected roadway situations, such as recurrent fog in combination with hazardous traffic patterns.