Human Identification Using a Smartphone Motion Sensor and Gait Analysis.

This research focuses on the modeling of human gait in order to recognize a human by their walking style. Gait analysis is a very important parameter, especially for biometric systems and also for healthcare applications. In this research, human gait is captured by using only a single Inertial Measurement Unit (IMU) sensor, which is positioned on top of the ankle. MATLAB mobile application is used to access the IMU sensor. A small study of 30 participants was conducted and two databases are then developed with different features. The first database contains only IMU sensor data, while the second database contains information on the height and gender of each participant. Both databases are used to train and test four different machine learning algorithms and their performances are compared. The experimental results show that the acquired setup performs well for classification purposes. Results from both databases are comparable and provide sufficient accuracy in the recognition task.