Analysis of Motion Detection using Social Force Model

• Main Authors: Wan Nur Azhani, W. Samsudin, Kamarul Hawari, Ghazali, Mohd Falfazli, Mat Jusof
• Format: Conference or Workshop Item
• Language: English
• Published: 2013
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://umpir.ump.edu.my/id/eprint/5038/1/fkee-2013-azhani-AnalysisOfMotion.pdf; http://umpir.ump.edu.my/id/eprint/5038/; http://worldconferences.net/proceedings/aics2013/toc/papers_aics2013/A062%20-%20WAN%20NUR%20AZHANI%20-%20ANALYSIS%20OF%20MOTION%20DETECTION%20USING%20SOCIAL%20FORCE%20MODEL.pdf
• Institution: Universiti Malaysia Pahang Al-Sultan Abdullah

Crowd behaviour detection is becoming a significant research topic in surveillance system in public places. This paper presents a method for the detection of abnormality in crowded scenes based on Social Force Model. For this purpose, Horn-Schunck optical flow is used in order to find the flow vector for all video frames. Using the vectors from this method, the interaction forces for each particle in video frames is calculated based on Social Force Model algorithm. The abnormal and normal frames are then classified by using a bag of words approach, whereby the region of anomalies in the abnormal frames are localized using interaction forces obtained in the previous experiment.