Combinational subsequence matching for human identification from general actions

Abstract

Except for gait analysis in a controlled environment, few have considered the use of motion characteristics for human identification, due to the complexity caused by the spatial nonrigidity and temporal randomness of human action. This work is a new attempt at mining biometric information from more general actions. A novel method for calculating the distance between two time series is proposed, where automatic segmentation and matching are conducted simultaneously. Given a query sequence, our method can efficiently match it against the gallery dataset. Local continuity and global optimality are both considered. The matching algorithm is efficiently solved by Linear Programming (LP). Synthetic data sequences and challenging broadcast sports videos are used to validate the effectiveness of our algorithm. The results show that action-based biometrics are promising for human identification, and the proposed approach is effective for this application. © 2013 Springer-Verlag.