Human action recognition based on kinect

被引：0

作者：

Yang, Mingya ^{[1
]}

Lin, Zhanjian ^{[1
]}

Tang, Weiwei ^{[1
]}

Zheng, Lingxiang ^{[1
]}

Zhou, Jianyang ^{[1
]}

机构：

[1] School of Information Science and Engineering

来源：

Journal of Computational Information Systems | 2014年 / 10卷 / 12期

关键词：

Action recognition; Human posture; Kinect;

D O I：

10.12733/jcis10733

中图分类号：

X9 [安全科学];

学科分类号：

0837 ;

摘要：

Human action recognition turns more important for the need of several applications such as personal assistive robotics and video surveillance. In this paper, we provide an effective method for human action recognition with angles of body parts and body motion vectors. We represent an human action as a sequence of body postures. Each body posture is represent by the body parts angles which are computed from the kinect outputted 3D skeletal joints. A motion vector is calculated from the sequence of body postures. We then use the state machine which generated from training data set to recognize the action from the body posture sequence and the corresponding motion vectors. The experimental results show that our method is real-time and has higher action recognition accuracy than the compared method in action recognition. It indicates that our method is efficient and effective to recognize human actions. What is more, our method has the advantage of view invariance and body size invariance. 1553-9105/Copyright © 2014 Binary Information Press.

引用

页码：5347 / 5354

页数：7

共 16 条

[1]

Pollack M.E., Autominder: An intelligent cognitive orthotic system for people with memory impairment, Robotics and Autonomous Systems, 44, 3-4, pp. 273-282

[2]

Yin J., Human action recognition in daily living, Journal of Computational Information Systems, 7, 100, pp. 3599-3605, (2011)

[3]

Raptis M., Real-time classification of dance gestures from skeleton animation, ACM SCA '11, pp. 147-156, (2011)

[4]

Sheng H., Human action recognition using star skeleton, VSSN '06 Proceedings of the 4th ACM International Workshop on Video Surveillance and Sensor Networks, pp. 171-178

[5]

Hsieh C.-H., Human action recognition using silhouette histogram, ACSC '11 Proceedings of the Thirty-Fourth Australasian Computer Science Conference, 113, pp. 11-16

[6]

Kellokumpu V., Human activity recognition using sequences of postures, Proc IAPR Conf. Machine Vision Applications, 3, pp. 570-573, (2005)

[7]

Wang L., Human action recognition from boosted pose estimation, Digital Image Computing: Techniques and Applications (DICTA), 2010 International Conference on, (2010)

[8]

Yamato J., Recognizing human action in time-sequential images using hidden markov model, computer vision and pattern recognition, Proceedings CVPR '92., 1992 IEEE Computer Society Conference on, (1992)

[9]

Zhang C., RGB-D Camera-based Daily Living Activity Recognition Journal of Computer Vision and Image Processing, 2-12, 4, (2012)

[10]

Lu X., View invariant human action recognition using histograms of 3D joints, Computer Vision and Pattern Recognition Workshops (CVPRW), (2012)

← 1 2 →