People identification based on dynamic determinants of human gait

In the paper a way of people identification, based on ground reaction forces during gait, is presented. The authors established that each individual has an unique gait pattern that can be described by quantitative parameters, calculated using measurements coming from the force plates. Fifteen volunteers took part in this study. Each person walked barefoot at least 100 times through 10-m-long walkway with the force plates built in. Determinants were calculated based on vertical and anterior-posterior components of the ground reaction force. The obtained parameters were used as an input matrix of the artificial neural network designed for identification of each person. Effectiveness of the recognition was assessed as root mean square error between expected and obtained output values. It was proved, that human identification based on presented determinants of the gait and artificial neural network is possible at a high level.