A real-time vehicle tracking method is proposed for trattlC monitoring system at roau mte^cc- tions, and the vehicle tracking module consists of an initialization stage and a tracking stage. Li- cense plate location b...A real-time vehicle tracking method is proposed for trattlC monitoring system at roau mte^cc- tions, and the vehicle tracking module consists of an initialization stage and a tracking stage. Li- cense plate location based on edge density and color analysis is used to detect the license plate re- gion for tracking initialization. In the tracking stage, covariance matching is employed to track the license plate. Genetic algorithm is used to reduce the computational cost. Real-time image tracking of multi-lane vehicles is achieved. In the experiment, test videos are recorded in advance by record- ers of actual E-police systems erage false detection rate and at several different city intersections. In the tracking module, the av- missed plates rate are 1.19%, and 1.72%, respectively.展开更多
Vehicle reidentification is an elegant solution for gathering several pieces of valuable traffic information, e.g., space mean speed, travel time, vehicle tracking, and origin/destination data. Recently, a number of v...Vehicle reidentification is an elegant solution for gathering several pieces of valuable traffic information, e.g., space mean speed, travel time, vehicle tracking, and origin/destination data. Recently, a number of vehiclereidentification algorithms utilizing inductive loop signals have been proposed to take advantage of the widespread availability of loop detectors. These algorithms, however, all directly utilize the raw inductance signals for pattern matching and feature extraction without deconvolution. The raw loop signals are essentially a convolved output between the true vehicle inductance signature and the loop system function, and thus a deconvolution is needed in order to expose the detailed features of individual vehicles. The purpose of this paper is to present a recent investigation on restoration of true inductance signatures by applying a blind deconvolution process. The main advantage of blind deconvolution over the conventional deconvolution is that the computation does not require modeling of a precise loop-detector system function. Experimental results show that the proposed blind deconvolution reveals much more detailed features of inductance signals and, as a result, increases the vehicle reidentification accuracy.展开更多
基金Supported by the National Natural Science Foundation of China(No.61005034)China Postdoctoral Science Foundation and under Grant(No.2012M510768)the Science Foundation of Hebei Province under Grant(No.F2012203182)
文摘A real-time vehicle tracking method is proposed for trattlC monitoring system at roau mte^cc- tions, and the vehicle tracking module consists of an initialization stage and a tracking stage. Li- cense plate location based on edge density and color analysis is used to detect the license plate re- gion for tracking initialization. In the tracking stage, covariance matching is employed to track the license plate. Genetic algorithm is used to reduce the computational cost. Real-time image tracking of multi-lane vehicles is achieved. In the experiment, test videos are recorded in advance by record- ers of actual E-police systems erage false detection rate and at several different city intersections. In the tracking module, the av- missed plates rate are 1.19%, and 1.72%, respectively.
文摘Vehicle reidentification is an elegant solution for gathering several pieces of valuable traffic information, e.g., space mean speed, travel time, vehicle tracking, and origin/destination data. Recently, a number of vehiclereidentification algorithms utilizing inductive loop signals have been proposed to take advantage of the widespread availability of loop detectors. These algorithms, however, all directly utilize the raw inductance signals for pattern matching and feature extraction without deconvolution. The raw loop signals are essentially a convolved output between the true vehicle inductance signature and the loop system function, and thus a deconvolution is needed in order to expose the detailed features of individual vehicles. The purpose of this paper is to present a recent investigation on restoration of true inductance signatures by applying a blind deconvolution process. The main advantage of blind deconvolution over the conventional deconvolution is that the computation does not require modeling of a precise loop-detector system function. Experimental results show that the proposed blind deconvolution reveals much more detailed features of inductance signals and, as a result, increases the vehicle reidentification accuracy.
