Higher-order spectral analysis of human motion

We describe a higher-order spectral analysis-based approach for detecting people by recognizing human motion such as walking or running. The periodic attribute of human motion lends itself to efficient spectral inspection. In the proposed method, the stride length is determined in every frame as the image sequence evolves. The bispectrum which is the Fourier transform of the triple correlation is a robust indicator of presence of periodicity. Triple correlation is robust as it is immune to any symmetrically distributed noise. The method is successfully tested on real video sequences.

[1]  C. L. Nikias,et al.  Signal processing with higher-order spectra , 1993, IEEE Signal Processing Magazine.

[2]  Larry S. Davis,et al.  Real-time periodic motion detection, analysis, and applications , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[3]  Karl Rohr,et al.  Incremental recognition of pedestrians from image sequences , 1993, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[4]  Yee-Hong Yang,et al.  Human body motion segmentation in a complex scene , 1987, Pattern Recognit..

[5]  Hironobu Fujiyoshi,et al.  Real-time human motion analysis by image skeletonization , 1998, Proceedings Fourth IEEE Workshop on Applications of Computer Vision. WACV'98 (Cat. No.98EX201).

[6]  Fang Liu,et al.  Finding periodicity in space and time , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[7]  Christian Wöhler,et al.  Motion-based recognition of pedestrians , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[8]  A. Lohmann,et al.  Triple correlations , 1984, Proceedings of the IEEE.

[9]  Mubarak Shah,et al.  Cyclic motion detection for motion based recognition , 1994, Pattern Recognit..

[10]  Rama Chellappa,et al.  Detection of people in images , 1999, IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339).