Skip to content
Research Article Open access CC BY 4.0

Human Motion Analysis from Depth Video Sequences Using Multi-scale and Multi-directional Features

Md. Saifuddin Tarafder, Md. Shahadat Hossain, Md. Rafiqul Islam

Current Journal of Applied Science and Technology · pp. 1–12 · Published 1 Apr 2017

10.9734/BJAST/2017/31153

Abstract

The emerging cost-effective depth sensors have made easier the action recognition task significantly. In this paper, we propose an effective method to analysis human actions from depth video sequences based on multi-scaling and multi-directional transformation which provide additional body shape and motion information for action recognition. In our method, corresponding to the front, side and top projection views, we generate three Depth Motion Maps (DMMs) over the entire video sequences. More specially, the multi-scaling and multi-directional transformations are implemented on the generated DMMs of a depth video sequence. Finally, the concatenation of these features is used as a feature descriptor for the depth video sequence. With this new feature descriptor, the l2-regularized collaborative representation classifier (l2- CRC) is utilized to recognize human actions. The recognition results of Microsoft Research (MSR) Action3D dataset show that our method significantly outperforms than the other existing methods, although our representation is much more compact.

3D Action Recognition depth maps depth motion maps Shearlet Transform MSR Action3D l2-CRC

References (20)

  1. 1 Real-time human pose recognition in parts from single depth images [DOI]
  2. 2 Action Recognition with Improved Trajectories [DOI]
  3. 3 A survey on vision-based human action recognition [DOI]
  4. 4 Human activity analysis [DOI]
  5. 5 Sparse Representation for Computer Vision and Pattern Recognition [DOI]
  6. 6 Action recognition based on a bag of 3D points [DOI]
  7. 7 Tikhonov Regularization and Total Least Squares [DOI]
  8. 8 HON4D: Histogram of Oriented 4D Normals for Activity Recognition from Depth Sequences [DOI]
  9. 9 Recognizing actions using depth motion maps-based histograms of oriented gradients [DOI]
  10. 10 Robust 3D Action Recognition with Random Occupancy Patterns [DOI]
  11. 11 Gait recognition using static, activity-specific parameters [DOI]
  12. 12 Real-time human action recognition based on depth motion maps [DOI]
  13. 13 STOP: Space-Time Occupancy Patterns for 3D Action Recognition from Depth Map Sequences [DOI]
  14. 14 Action Recognition from Depth Sequences Using Depth Motion Maps-Based Local Binary Patterns [DOI]
  15. 15 Evolutionary joint selection to improve human action recognition with RGB-D devices [DOI]
  16. 16 Human action recognition with skeletal information from depth camera [DOI]
  17. 17 Real-Time Human Action Recognition Using DMMs-Based LBP and EOH Features [DOI]
  18. 18 Global versus structured interpretation of motion: moving light displays [DOI]
  19. 19 An Update-Describe Approach for Human Action Recognition in Surveillance Video [DOI]
  20. 20 Digital PIV: A Challenge for Feature Based Tracking

Cited by 0

No indexed citations yet.

Article metrics

Real usage data collected on this platform.

61

Page views

0

PDF downloads

39

Outbound clicks

0

Citations

Views over time

Views by country

Approximate, from request IP at view time — not citizenship or institution. Countries with fewer than 5 views are grouped as "Other".

Traffic sources

Referring site, by host.

Views and downloads exclude known bots/crawlers. Citations combines this platform's own DOI-resolved index with each external source's own reported total — see Cited by above for individually listed citing works. Last refreshed 0 seconds ago.