Virtual S.L.A.M Dunk!

Simultaneous Localization and Mapping (S.L.A.M.) and Tracking

There are a few measurements that are important in to XR movement sensed by devices, such as Tracking, Calibration, and Registration.

• Tracking is the sensing and measuring of a XR device’s position and orientation. In order for a 3D digital object to appear in physical space we need to know the position and orientation of the device to the object it is tracking; in other words, we need to know the proximity to real-life object.
• Calibration is the measurement between two devices (reference and calibrated objects). It uses a geometric measurement reference point known as coordinate systems to aid in location when tracking.
• Registration is the unification of coordinates between digital 3D objects and real-life physical objects; it requires a camera and head movement. It can be interrupted by latency and update rate, making the virtual reality produce lag or a glitchy experience.

Device Sensors

IMU’s (inertial measurement units) are the device’s multiple sensors that are important when tracking. These sensors are magnetometers, gyroscopes, accelerometers, as well as optical sensors (cameras). Magnetometers act as an electronic compass. It measures the direction of the Earth’s magnetic field and determines true North; known as ‘global orientation’.
Gyroscopes measures rotation vibration or velocity of a device, aka when the device is rotated the Image is reoriented. Accelerometers detects positioning by measuring movements; it also aids in counting steps. Optical tracking uses sensor in the camera. Cameras have a Bayer layer which senses light and translates it into an image. Same with optical tracking, it uses the Bayer layer in device’s camera to sense light and create and image, which can essentially be tracked.

Model-Based Vs. Model-Free Tacking

Model transformation is the relationship between 3D local object and world coordinates. 3D local objects can be static (at a fixed position or ‘local’) …or move about the physical environment. 3D objects are measured and controlled by the device. World coordinates help track an object when it is moving. It uses geometric data to calculate and calibrate its location. To measure and track larger space we need to use GPS (Global Positioning System) which uses satellites to produce the coordinates to a devices location. It is noted, Stationary trackers are easier to use and build, because of less uncontrolled variables and sensor errors to deal with. Overall, the orientation of the user is measured by the sensors within the device. GPS measures positioning of a user in real time. (AR: Principles and Practice_Chapter 3 Tracking)

To summarize: Model-Based tracking is a static or fixed object acquired before tracking; it only uses 3 DOF. Model-Free tracking requires more flexibility because the model or object is acquired or tracked on the fly, using 6DOF. This is known as Simultaneous Location and Mapping (S.L.A.M.). This combines 3D tracking and GPS scanning of location. (AR: Principles and Practice_Chapter 3 Tracking)

For more of an understanding on S.L.A.M. Click Here for more reading, or Click Here to watch a video for more clarification.

Positional Tacking

Positional tacking estimates the position relative to the environment around it. It uses 6 degrees of freedom; head movement and body positioning. This is where SLAM comes in. It uses sensors in the device to help map out unknown the surrounding and physical environments. This relies not only on sensors in the device but GPS tracking as well.

For more on Postional tracking Click Here.

All Images are copyright free from Unsplash.com