Open Access Open Access  Restricted Access Subscription or Fee Access

Automated Complex Activity Recognition in Multiple Person Interaction

J. Arunnehru, M. Kalaiselvi Geetha

Abstract



Vision-based activity recognition is an essential and challenging theme in computer vision community, with various significant applications including smart video surveillance, abnormal detection system and human - computer interactions. In this paper, an activity recognition
method Cumulative Difference Intensity Distance Group Pattern (CDIDGP) based on motion information extracted from n - cumulative frame difference images. Initially, motion information is obtained by performing absolute difference between two successive frames, then difference
frame is accumulated through n - frames forming an n - cumulative frame difference images. The CDIDGP features extracted from the n - frame difference images characterize the motion vector information of the activity. The experiments are carried out using UT-Interaction dataset
(Set 1 and Set 2) consider only full-body movements of interaction between two persons and the extracted CDIDGP feature set are modeled by the various tree based classifier like Random Forest, Decision Tree (J48) and Naive Bayes. Experimental results show that this approach is effective and efficient to recognize interaction activity with an overall recognition accuracy of 90.28% and 86.13% on the UT-Interaction dataset (Set 1 and Set 2) respectively. The performance of the proposed approach outperforms previously published approaches.

Keywords


Human Activity Recognition, Feature Extraction, Temporal Difference, Motion Analysis, Random Forest, Video Surveillance.

Full Text:

PDF

Refbacks

  • There are currently no refbacks.


Disclaimer/Regarding indexing issue:

We have provided the online access of all issues and papers to the indexing agencies (as given on journal web site). It’s depend on indexing agencies when, how and what manner they can index or not. Hence, we like to inform that on the basis of earlier indexing, we can’t predict the today or future indexing policy of third party (i.e. indexing agencies) as they have right to discontinue any journal at any time without prior information to the journal. So, please neither sends any question nor expects any answer from us on the behalf of third party i.e. indexing agencies.Hence, we will not issue any certificate or letter for indexing issue. Our role is just to provide the online access to them. So we do properly this and one can visit indexing agencies website to get the authentic information.