Title :
Behavior recognition based on dynamic programming and Concurrence Probabilistic Petri Nets
Author_Institution :
Sapientia Hungarian Univ. of Transylvania, Romania
Abstract :
In this paper, we present an approach based on dynamic programming and Concurrence Probabilistic Petri Nets for recognition and matching human action and behavior. Each human motion is represented by the body parts angular variation. Each body part angular motion is represented by one-dimensional time series. These time series are compared separately for every body part with templates, using dynamic programming (DTW). The results of the comparisons are used as input for the Concurrence Probabilistic Petri Nets that classifies the human action and behavior.
Keywords :
Petri nets; behavioural sciences; dynamic programming; image matching; image motion analysis; probability; time series; behavior recognition; body part angular motion; body parts angular variation; concurrence probabilistic Petri net; dynamic programming; human action matching; human action recognition; human motion; one-dimensional time series; Artificial neural networks; Hidden Markov models; Humans; Petri nets; Probabilistic logic; Time series analysis; Tracking;
Conference_Titel :
Intelligent Computer Communication and Processing (ICCP), 2010 IEEE International Conference on
Conference_Location :
Cluj-Napoca
Print_ISBN :
978-1-4244-8228-3
Electronic_ISBN :
978-1-4244-8230-6
DOI :
10.1109/ICCP.2010.5606443
