Joint movement similarities for robust 3D action recognition using skeletal data

Document Type

Journal Article

Publisher

Academic Press Inc.

School

School of Computer and Security Science

RAS ID

21075

Comments

Pazhoumand-Dar, H., Lam, C.-P., Masek, M. (2015). Joint movement similarities for robust 3D action recognition using skeletal data. In Journal of Visual Communication and Image Representation, 30, 10-21.Available here.

Abstract

Abstract Human action analysis based on 3D imaging is an emerging topic. This paper presents an approach for the problem of action recognition using information from a number of action descriptors calculated from a skeleton fitted to the body of a tracked subject. In the proposed approach, a novel technique that automatically determines discriminative sequences of relative joint positions for each action class is employed. In addition, we use an extended formulation of the longest common subsequence algorithm as a similarity function, which allows the classifier to reliably find the best match for extracted features from noisy skeletal data. The proposed approach is evaluated using two existing datasets from the literature, one captured using a Microsoft Kinect camera and the other using a motion capture system. The experimental results show that the approach outperforms existing skeleton-based algorithms in terms of its classification accuracy and is more robust in the presence of noise when compared to the dynamic time warping algorithm for human action recognition.

Abstract Human action analysis based on 3D imaging is an emerging topic. This paper presents an approach for the problem of action recognition using information from a number of action descriptors calculated from a skeleton fitted to the body of a tracked subject. In the proposed approach, a novel technique that automatically determines discriminative sequences of relative joint positions for each action class is employed. In addition, we use an extended formulation of the longest common subsequence algorithm as a similarity function, which allows the classifier to reliably find the best match for extracted features from noisy skeletal data. The proposed approach is evaluated using two existing datasets from the literature, one captured using a Microsoft Kinect camera and the other using a motion capture system. The experimental results show that the approach outperforms existing skeleton-based algorithms in terms of its classification accuracy and is more robust in the presence of noise when compared to the dynamic time warping algorithm for human action recognition. © 2015 Elsevier Inc. All rights reserved.

DOI

10.1016/j.jvcir.2015.03.002

Access Rights

subscription content

Share

 
COinS