Study On Vehicle Detection And Tracking Method Based On Video Sequences In ITS

This thesis mainly discusses the fundamental theories and key technologies of vehicle detection and tracking system based on video sequences in ITS (Intelligent Transportation System). Traffic detection and information collection has become an important subject in ITS, and vehicle detection, shadow elimination, vehicle recognition and tracking are the basic parts. Based on summarizing and analyzing the existing technologies of vehicle detection and tracking system, this thesis studies these issues, and brings forward new methods. Moreover, experiments are implemented to demonstrate the validity. The main contents of the study include such aspects as following:1. Study on foreground extractionThis thesis proposes a new foreground extraction algorithm based on edge information. First extract accurate background edge of current frame by background edge extracting model, and eliminate it. Then using morphological method to eliminate noise, and ascertain the foreground regions to divide up vehicles. The contrast experiment with background subtraction algorithm shows that this algorithm is simple and efficient, it is suitable for complicated traffic scene of city zone, it can figure out camera lens shaking and roadside trees shaking problems. This algorithm is better than background subtraction algorithm.2. Study on shadow eliminationTo solve the shadow problem, this paper develops an effective shadow distinguishing and eliminating method based on intensity information, edge information and transcendental knowledge. Firstly detect shadow of single tint vehicle by intensity and edge information, then memorize the ubiety between shadow and vehicle to update transcendental knowledge. When several cars moving together or fuscous vehicle come in, detect shadows combined with intensity information, edge information and transcendental knowledge. Experimental results demonstrate that the proposed method can eliminate the shadow accurately.3. Study on vehicle trackingAiming at the particularity of vehicle moving in video sequences, this thesis proposes a new vehicle tracking algorithm,defines a moving vector between frames, using the vehicle rate and acceleration in current frame to prognosticate the rate in next frame, and match them by combining with shape information of the tracking regions. The experimental results show that this algorithm is simple and efficient, it is suitable for the real-time vehicle detection and tracking system.