FILIP KOLASA / paper / bsc-thesis
/paper/bsc-thesis

Motion Capture Technology Using Inertial Sensors

This paper examines the principles and applications of Inertial Measurement Units (IMUs), with a focus on inertial sensors and data fusion techniques used for orientation and motion estimation. The work highlights inertial motion capture systems as a representative application, demonstrating the use of distributed IMU networks for human motion reconstruction across scientific, industrial, and applied domains.

Status
published
Version
v1.0
Last updated
2024-01-15

Abstract

Inertial Measurement Units (IMUs) constitute a particularly significant area of study. When combined with appropriate data fusion algorithms, they enable relatively accurate estimation of system orientation and, in certain cases, position. When integrated with external sources of information such as GPS, whose role is to correct the estimated position of the unit based on approximately ground-truth data, IMUs make it possible to construct precise navigation systems applied, for example, in the military sector.

This engineering thesis addresses issues related to Inertial Measurement Units through the analysis of their key components—namely inertial sensors. These sensors, in the form of accelerometers and gyroscopes, often complemented by magnetometers, function as integrated nine-degree-of-freedom systems responsible for tasks such as gesture detection in smartwatches or automatic rotation of the displayed image on smartphone screens depending on device orientation.

The aforementioned examples, however, illustrate the use of Inertial Measurement Units primarily as auxiliary functionalities within larger devices. An application that more fully demonstrates the capabilities of this technology is an inertial motion capture system. Such a system constitutes a distributed network of interconnected and cooperating Inertial Measurement Units that collect motion data from individual segments of the human skeletal structure in order to enable subsequent reconstruction of movement based on recorded measurements. Inertial motion capture systems are applied in various domains, including the entertainment industry for computer animation production, the medical field for detection and diagnosis of postural disorders, sports science for motion analysis and identification of elements requiring improvement, virtual and augmented reality systems to enhance user experience, industrial applications for remote control of humanoid robots or similarly controlled platforms, and training environments that employ virtual simulations for personnel education.