A Comparative Analysis of Various Representations of Human Action Recognition in a Video
|Akila.K1, Chitrakala.S 2
|Related article at Pubmed, Scholar Google|
The action recognition is an automated analysis of ongoing events and their context from video data. The process of recognizing and understanding of human actions from videos still remains a challenging problem due to the large variations in human appearance, posture and body size within the same class; hence a pattern recognizer needs to be built into the video system. Robust solutions to this problem have applications in domains such as visual surveillance, video retrieval and human–computer interaction. Various features are used to recognize humans and they are categorized into four groups based on the visual representations such as spatio-temporal based visual representation, shape or pose based visual representation, interest point based visual representation and motion or optical flow based visual representation. The performance of these works is analyzed efficiently with the evaluation metrics such as Precision, Recall, F-Measure and Accuracy. It is tested in both controlled and uncontrolled environments. This paper helps to develop novel technologies over the biometric recognition system.