“3D reconstruction of walking behaviour using a single camera” by Somasundaram and Parent

  • ©Arunachalam Somasundaram and Richard (Rick) E. Parent

Conference:


Type(s):


Interest Area:


    Application

Title:

    3D reconstruction of walking behaviour using a single camera

Session/Category Title:   Human Figure Animation


Presenter(s)/Author(s):



Abstract:


    A consumer-grade camera is used to view the motion of individuals moving in an area, such as that observed in the lobby of a building. By assuming individuals are engaged in fairly normal walking behavior, we can extract sufficient information to synthetically reconstruct the scene using simple image processing techniques. The reconstructed scene can then be viewed from any angle and used to track interesting individuals. Features of the person such as shirt and pants color can be extracted from the video and applied to the synthetic model thus allowing the real individual to be recognized from viewing the reconstruction. The technique proposed has the advantage of using a single camera and would find applications in gait analysis and security. This is a work in progress and we present some initial results of a single moving figure.

References:


    1. Tsai, R. Y. 1987. A versatile camera calibration technique for high accuracy 3D machine vision metrology using off-the shelf tv cameras and lenses. IEEE JOURNAL OF ROBOTICS AND AUTOMATON RA – 3, 4, 323–344.


ACM Digital Library Publication:



Overview Page: