1. Description: this video demonstrates the feasibility of simultaneously tracking interacting pedestrians, bicycles, motorcycles, cars and trucks in crowded urban areas. In this video, red points denote scan points belonging to stationary objects, blue points denote scan points belonging to moving objects, and yellow points denote scan points belonging to new objects which cannot be classified. White rectangles denote the virtual measurements of the interaction models, green rectangles denote segments of moving objects and purple lines denote trajectories of tracked objects. Anomalous event detection is shown with red rectangles.

2. Player Information: Windows Media Player 9.0+ 

3. Contact Information: Chieh-Chih (Bob) Wang, Department of Computer Science and Information Engineering, National Taiwan University. Tel: +886-2-33664888 ext 522. Email: bobwang@ntu.edu.tw


Attachment filename: 0994_VI.mpg (5217304 Bytes)