A Comprehensive Study on Human Activity Recognition

Recognizing the activity of a person and his motive from video sequences and sensor data is one of the major challenges in human-computer interaction and computer vision. Identifying the activities, processing them for classification and making decision on whether it is walk, sit, stand or fall is the prime functionality of Human Activity Recognition (HAR). Elderly people on been alone at home, face many problems including collapsing, vomiting, chest pain, stomach ache etc. Therefore, elderly people or patients could be definitely supported by HAR which would help to monitor their activity and any changes in their behavior or any occurring critical activity. This review provides a comprehensive study on the state-of-the-art HAR approaches along with the datasets used. The performance metrics used for the experimental evaluations are also analyzed and highlighted. The motivation of the research on HAR and the directions for future research are also explored.