Walking motion recognition system by estimating position and pose of leg mounted camera device using visual SLAM

Abstract

This paper presents the method for recognizing the walking motion of the user with estimating position and pose of leg mounted camera devices by employing Visual SLAM. Approximate displacement of the position and orientation of the sensor terminal can be estimated using output values of the acceleration sensor, the gyro sensor, and the geomagnetic sensor. It is also considered that the position and orientation of the terminal can be estimated by the image captured by the camera mounted on each terminal. The system records surrounding scenery of the user with each camera device attached to each part of the user's leg, and estimates the 3D position and orientation of each camera device by using Visual SLAM. Finally, we develop the walking training support system that does not require a special environment.