Analysis of Foot-pressure Data to Classify Mobility Pattern

The pressure exerted by foot, while a person is standing still for a while or moving or doing any physical activity, is a rich source of information. The continuous signal obtained throughout the day, collected by pressure-sensors on shoe sole, could be analyzed to obtain simple to complex facts about the person’s health conditions and habits. It could be used to measure body-weight and balance, while the person is standing. It could as well be used to find the total calorie burnt during movement activities throughout the day. Varied applications would need different number of sensors spread over inside or outside the shoe-sole. In this work, we restrict our investigation to simple applications like, measuring the body weight when the person is standing still, or the speed when the person is moving, or whether she/he is climbing up or down the stairs. Our aim is to use as few sensors as possible, and the algorithm simple and efficient. For measuring body-weight and movement speed, we could achieve nearly 100% accuracy. We could also classify between climbing up or down the stairs with 100% accuracy. All these could be accomplished by a single or a pair of sensors. It is also revealed that the optimum location of the sensor/s for the highest accuracy varies from person to person.