Đang chuẩn bị liên kết để tải về tài liệu:
Human action recognition using dynamic time warping and voting algorithm

This paper presents a human action recognition method using dynamic time warping and voting algorithms on 3D human skeletal models. In this method human actions, which are the combinations of multiple body part movements, are described by feature matrices concerning both spatial and temporal domains. | VNU Journal of Science: Comp. Science & Com. Eng., Vol. 30, No. 3 (2014) 22-30 Human Action Recognition Using Dynamic Time Warping and Voting Algorithm(1) Pham Chinh Huu*, Le Quoc Khanh, Le Thanh Ha Faculty of Information Technology, VNU University of Engineering and Technology 144 Xuan Thuy, Cau Giay, Hanoi, Vietnam Abstract This paper presents a human action recognition method using dynamic time warping and voting algorithms on 3D human skeletal models. In this method human actions, which are the combinations of multiple body part movements, are described by feature matrices concerning both spatial and temporal domains. The feature matrices are created based on the spatial selection of relative angles between body parts in time series. Then, action recognition is done by applying a classifier which is the combination of dynamic time warping (DTW) and a voting algorithm to the feature matrices. Experimental results show that the performance of our action recognition method obtains high recognition accuracy at reliable computation speed and can be applied in real time human action recognition systems. © 2014 Published by VNU Journal of Science. Manuscript communication: received 10 December 2013, revised 04 March 2014, accepted 26 March 2014 Corresponding author: Pham Chinh Huu, phamchinhhuu@gmail.com Keywords: Human action recognition, feature extraction, Dynamic time warping. e 1. Introduction1 Human action recognition has become an interesting computer vision research topic for the last two decades. It is motivated by a wide range of potential applications related to video surveillance, human computer interaction aimed at identifying an individual through their actions. The evaluation of human behavior patterns in different environments has been a problem studied in social and cognitive sciences. However, it is raised as a challenging approach to computer science due to the complexity of data extraction and its analysis. _ 1 This work was supported by