Human Interactions Recognition using Bag of Words

R.Newlinshebiah, S.P.Sivasubbu, V.Sivasankar

A video surveillance system can be defined as a technological tool that assists humans by providing an extended perception and capability of capturing interesting activities in the monitored scene. This paper describes a methodology for automated recognition of one to one human interactions such as handshake, kicking and hugging. The frame work consists of background subtraction followed by feature extraction (Speed Up Robust Features) and action classification using SVM classifier. It is computationally efficient and invariant to occlusion, lightning. The method produces good categorization accuracy and precision. Human behaviour recognition has various applications such as human-computer interfaces; content based video retrieval, Visual monitoring & surveillance.