Stride and Cadence as a Biometric in Automatic Person Identification and Verification
FGR '02 Proceedings of the Fifth IEEE International Conference on Automatic Face and Gesture Recognition
Learning Pedestrian Models for Silhouette Refinement
ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
The HumanID Gait Challenge Problem: Data Sets, Performance, and Analysis
IEEE Transactions on Pattern Analysis and Machine Intelligence
Individual Recognition Using Gait Energy Image
IEEE Transactions on Pattern Analysis and Machine Intelligence
A Layered Deformable Model for Gait Analysis
FGR '06 Proceedings of the 7th International Conference on Automatic Face and Gesture Recognition
Automatic Gait Recognition using Dynamic Variance Features
FGR '06 Proceedings of the 7th International Conference on Automatic Face and Gesture Recognition
Modelling the effects of walking speed on appearance-based gait recognition
CVPR'04 Proceedings of the 2004 IEEE computer society conference on Computer vision and pattern recognition
On automated model-based extraction and analysis of gait
FGR' 04 Proceedings of the Sixth IEEE international conference on Automatic face and gesture recognition
Fusion of static and dynamic body biometrics for gait recognition
IEEE Transactions on Circuits and Systems for Video Technology
| Hi-index | 0.00 |
Period detection and cycle partitioning are always the very beginning for most gait recognition algorithms. Badly segmented silhouettes and random fluctuations in walking speed are two of the main problems for this basic but important issue. In this paper, we propose a method of cycle partitioning that is adaptive to silhouette quality and speed fluctuations. To do that, autocorrelation on sliding window is proposed to quantify the silhouette quality into "trusted zones" and "uncertain zones". Prior period estimation and observation of fluctuations are incorporated to obtain more precise cycle detection. One criterion based on the difference of Common Phase Frames (CPF) is proposed to evaluate the precision of detection. In experiment, our method was compared with the traditional autocorrelation method using sequences from the USF gait database. The results showed the improved cycle partitioning performance of the proposed method.