Dynamic Sensor Matching based on Geomagnetic Inertial Navigation

TL;DR

The paper addresses the challenge of aligning data from multiple depth sensors in a common, GPS-unreliable environment. It proposes a magnetically defined world coordinate system and a geomagnetic inertial navigation pipeline that fuses IMU and magnetometer data to re-reference sensor positions. The main contributions are the global sensor matching concept anchored to Earth's magnetic field, a calibration/initiation and transformation workflow to establish the magnetically referenced WCS, and an experimental evaluation that demonstrates improved stability and alignment under varied conditions. This work enables real-time, GPS-free multi-sensor environment mapping with potential applications in indoor/outdoor depth sensing, while highlighting remaining sensitivity to external magnetic disturbances and the need for further filtering and longer-term validation.

Abstract

Optical sensors can capture dynamic environments and derive depth information in near real-time. The quality of these digital reconstructions is determined by factors like illumination, surface and texture conditions, sensing speed and other sensor characteristics as well as the sensor-object relations. Improvements can be obtained by using dynamically collected data from multiple sensors. However, matching the data from multiple sensors requires a shared world coordinate system. We present a concept for transferring multi-sensor data into a commonly referenced world coordinate system: the earth's magnetic field. The steady presence of our planetary magnetic field provides a reliable world coordinate system, which can serve as a reference for a position-defined reconstruction of dynamic environments. Our approach is evaluated using magnetic field sensors of the ZED 2 stereo camera from Stereolabs, which provides orientation relative to the North Pole similar to a compass. With the help of inertial measurement unit informations, each camera's position data can be transferred into the unified world coordinate system. Our evaluation reveals the level of quality possible using the earth magnetic field and allows a basis for dynamic and real-time-based applications of optical multi-sensors for environment detection.

Figures (9)

• Figure 1: Point Cloud Matching: The resulting point cloud (right) is matched from several depth images (top and bottom left). The arrows of $\lambda_\mathrm{1}$ and $\lambda_\mathrm{2}$ represent the camera orientation.
• Figure 2: Combination of 3D Depth Sensing and Smart Sensor Architecture: Relationships between the stereo camera and IMU integrated gyroscope and magnetometer in a defined world frame coordinate system. The received images of $C_\mathrm{R}$ and $C_\mathrm{L}$ are located in the extrinsic camera frame mue21. In contrast to the cameras, the IMU is originally located in body frame.
• Figure 3: Conceptual Representation of Global Sensor Matching: Description of the geometric relationships between the sensor points $\{ B_\mathrm{1}\}$ ,$\{B_\mathrm{n}\}$ and common origin $\{N_\mathrm{p}\}$ in the WCS.
• Figure 4: Calibration of the Magnetic Field Sensor: The magnetometer is integrated in the chassis of the ZED 2 from Stereolabs. Each colour represents a direction of rotation (red: $R_\mathrm{\phi}$, green: $R_\mathrm{\theta}$, blue: $R_\mathrm{\psi}$).
• Figure 5: Pipeline of Space Determination: The different steps of movement transformations in a WCS.