| Paper Details: | Downloads: 461 |
| Serial Number: | P1150535149
|
| Title: | Higher-Order Spectra Analysis of Silhouette for Gait Recognition
|
| Authors: | Yi Luo and Mehmet Celenk
|
| Abstract: | Recent advances in video-technology have generated an increasing demand for automated surveillance and human recognition systems based on different human biometrics such as face, gait, iris, hands, and palms. In this paper, we present a simple, yet effective gait recognition algorithm using the Radon transform and higher-order spectra (HOS) analysis. For gait recognition, a person’s silhouette in a sequence of images is detected by a least-median-of-squares (LMeDS) background modelling method. The silhouettes extracted from each frame in the image sequence are subjected to a statistical mean-shape generation process. The Radon transform and HOS analysis are utilized as an image-sequence processing tool to extract a given gait’s feature vector. This is then compared with the known gait signatures for a possible match by computing the Euclidean distances between the unknown gait signature and the available gait-classes’ representatives described in the Radon-projected HOS invariants. The gait in question is classified as belonging to one of the existing gait classes in the least Euclidean distance sense.
|
| Keywords: | Biometrics, gait recognition, higher-order spectra (HOS), Radon transform, principal component analysis (PCA), minimum-distance Euclidean classifier
|
| Journal/Conference: | ICGST Conference on Graphics, Vision and Image Processing, GVIP-05
|
| Volume: |
|
| Issue: |
|
| Submission Date: | 9/1/2005 12:00:00 AM
|
| Review Date: | 11/1/2005 12:00:00 AM
|
| Publishing Date: | 12/19/2005 12:00:00 AM
|
| Article Downloads: | 461
|
| Download: |
|