Comparing VIDEO CAPTURE and 3D IMU MOTION SYSTEMS
Using traditional, lab-based video to capture motion data delivers single-plane kinematics with very precise positional information based on markers. To determine velocity, the position signal must be differentiated. First, the derivative operation required to obtain velocity amplifies the small amount of noise in the original position information, resulting in large velocity errors, especially during fast motions. Secondly, the double-derivative operation required to obtain acceleration amplifies noise to such an extent that the derived acceleration values may not be useful.
In comparison, Noraxon’s 3D Motion Capture Inertial Measurement Unit (IMU) Systems take a different approach by providing multiplanar kinematics and rotation data. Instead of measuring marker position, the angular velocity of body segments are directly measured using each sensor’s on-board gyroscopes. To determine body position, this angular velocity signal is integrated. At the same time as this integration is taking place, the sensor’s on-board accelerometers are filtered in such a way as to accurately separate the [downward] gravity vector from the sensor’s linear acceleration vector. The gravity vector is then used to correct the drift accumulated during the gyroscope integration operation, while the linear acceleration is used as a (nearly direct) measurement of segment acceleration. With IMUs, acceleration, angular velocity and angular orientation measurements are all precise. Noraxon’s Notion Capture IMUs provide angular orientation estimates that are on par with video capture systems, and significantly higher accuracy for velocity and acceleration terms even during very rapid movements.
What are Inertial Measurement Units
An IMU is a Micro-Electro-Mechanical System (MEMS) typically comprised of 3 accelerometers and 3 gyroscopes. The more advanced systems also include 3 magnetometers. These tri-axial sensors are typically referred to as “3 Degrees of Freedom” (3 DoF) IMU sensors. 3 DoF means the sensors give only orientation (or tilt) data in 3 axes. The more advanced IMUs are capable of 6 DoF; these systems also provide translation or movement data indicating where the sensor is located in free space. Many IMU manufacturers misuse these terms and count the number of sensors as DoF, which is a misrepresentation of the inherent characteristics of IMUs. One example of this is when a sensor that includes 3 Gyroscopes, 3 Accelerometers and 3 Magnetometers is touted as a 9 DoF system.
When are IMUs the right choice for motion capture?
REAL TIME VISUALIZATION:
→ Inertial motion tracking systems offer real-time motion tracking visualization and requires little to no post-processing.
USE IN ANY ENVIRONMENT:
→ Inertial sensors can be used anywhere, inside or outside and with no restrictions on capture volume.
GUARANTEED DATA CAPTURE:
→ Inertial sensors do not suffer from any line-of-sight occlusions or data dropouts.
HIGH SAMPLING RATES:
→ IMU sensors have very high internal sampling rates, typically at least 800 Hz.
→ Inertial sensors directly measure linear accelerations and angular rate changes without the need for differentiating positional data. This allows for the direct estimation of force data.
What are some of the drawbacks of IMUs?
→ IMU systems with multiple sensors can be restrictive and inhibit the movements of the subject for certain activities. With the recent development of small wireless inertial measurement systems, inhibited movement is becoming less of a problem.
→ Global position data is not provided and must be calculated using either additional GPS sensors or with more advance systems that have a very accurate magnetometer combined with advanced Kalman filtering routines.
→ Positional data is typically not as accurate as either optical or magnetic tracking systems due to the mathmetical estimation process required.
→ Inertial sensors, due to the nature of the technology, usually have some level of positional drift over time.