Human Behaviour Recognition using Fuzzy System in Videos

Human behavior can be detected and analyzed using video sequence is a latest research topic in computer vision & machine learning. Human behavior is used as a basis for many modern applications, such as video surveillance, content-based information retrieval from videos etc. HBA (Human behaviour analysis) is tricky to design and develop due to uncertainty and ambiguity involved in people’s daily activities. To address this gap, we propose hierarchical structure combining TDNN, tracking algorithms, and fuzzy systems. As a result, HBA system performance will be improved in terms of robustness, effectiveness and scalability.