Gait Recognition on Features Fusion Using Kinect

  • Tianqi YangEmail author
  • Yifei Hu
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 905)


The paper proposes a gait recognition method which is about multi-features fusion using Kinect. The data of 3D skeletal coordinates is obtained by Kinect, and the multi-features are as follows. Firstly, the human skeletal structure is treated as a rod-shaped skeletal model, and it can be simple and convenient to reflect the static characteristics of structure of the human body from the overall. Secondly, the angle of the hip joint is observed during walking so that the dynamic characteristics of the gait information are reflected from the local area. Thirdly, the key gait body postures features are selected from a special gait to reflect the characteristics of walking. Then, the three gait features information is fused, which improves the overall recognition rate. After removing the noise from the bone data, in order to fully reflect the uniqueness of the individual, gait features are extracted from multiple angles, including both static and dynamic features. For finding the center of mass, the distance from the center of mass to the main joint point are calculated to measure the change in the center of mass, and the characteristics of hip joints reflecting the changes of the lower extremity joints while walking. The paper classify each feature separately by using the Dynamic Time Warping (DTW) and K-Nearest Neighbor (KNN) algorithm, which is used at the decision level. The experimental results show that the proposed method achieves a better recognition rate and has a good robustness.


Gait recognition Features fusion Kinect 



This study is supported by Science and Technology Project of Guangdong (2017A010101036).


  1. 1.
    Lee, T.K.M., Belkhatir, M., Sanei, S.: A comprehensive review of past and present vision-based techniques for gait recognition. Multimed. Tools Appl. 72(3), 2833–2869 (2014)CrossRefGoogle Scholar
  2. 2.
    Lee, L., Grimson, W.E.L.: Gait analysis for recognition and classification. In: Proceeding of Fifth IEEE International Conference on Automatic Face and Gesture Recognition, pp. 148–155 (2002)Google Scholar
  3. 3.
    Cunado, D., Nixon, M.S., Carter, J.N.: Automatic extraction and description of human gait models for recognition purposes. Comput. Vis. Image Underst. 90(3), 1–41 (2003)CrossRefGoogle Scholar
  4. 4.
    Yoo, J.H., Nixon, M.S., Harris, C.J.: Extracting gait signatures based on anatomical knowledge. In: Proceeding of The British Machine Vision Association and Society for Pattern Recognition, London (2002)Google Scholar
  5. 5.
    Preis, J., Kessel, M., Linnhoff-Popien, C.: Gait recognition with Kinect. In: 1st International Workshop on Kinect in Pervasive Computing (2012)Google Scholar
  6. 6.
    Kastaniotis, D., Theodorakopoulos, I., Theoharatos, C., Economou, G., Fotopoulos, S.: A framework for gait-based recognition using Kinect. Pattern Recognit. Lett. 68, 327–335 (2015)CrossRefGoogle Scholar
  7. 7.
    Ahmed, F., Paul, P.P., Gavrilova, M.L.: DTW-based kernel and rank-level fusion for 3D gait recognition using Kinect. Vis. Comput. Int. J. Comput. Graph. 31(6–8), 915–924 (2015)Google Scholar
  8. 8.
    Luo, Z.P., Yang, T.Q.: Gait recognition by decomposing optical flow components. Comput. Sci. 43(9), 295–300 (2016)Google Scholar
  9. 9.
    Chen, X., Yang, T.Q.: Cross-view gait recognition based on human walking trajectory. J. Vis. Commun. Image Represent. 25, 1842–1855 (2014)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  1. 1.Jinan UniversityGuangzhouChina

Personalised recommendations