Biometric User Identification with Dynamic Footprint

This paper presents a biometric user identification method based on user´s gait. We divide obtainable features from user´s gait into two categories: walking pattern and stepping pattern (dynamic footprints), and this paper considers an approach of identifying user with dynamic footprints. In this study, we gather gait samples from 10 users while walking over the UbiFloorII. A software module is developed to extract dynamic footprints from the samples acquired, and PCA (principal component analysis) and neural network technique are employed to identify the user with extracted features. The experimental results show 90% recognition accuracy. This approach shows the feasibility of our deployed features or dynamic footprint acquired from the UbiFloorII as biometric identifiers.