Fusion of visual tracking and inertial measurement units for movement analysis

Abstract

The study analyzed the possibility of combining visual tracking and data inertial measurement units (IMU). The block diagram of a motion registration system based on a combination of visual tracking data and IMU data is proposed. Optical data from cameras (at least two) are used to calculate the position. In parallel, IMU data is also used to calculate the position. All this information is combined in the appropriate block. Specialized probability filters should be used to fuse dissimilar data. After analyzing the available mathematical methods, the most optimal is the use of an extended Kalman filter