We propose the use of a new paradigm for motion tracking technology that is optically based, however instead of using high speed megapixel cameras (photo-detector array) to capture features or tags (illuminated or self-illuminating photo-emitters) spread in the environment, the capture process is inverted by spreading hundreds or thousands of single pixel cameras (markers with single photo-detector and on board processing) that can detect very high-speed (hundreds of thousands of times per second) binary patterns generated by illuminating projectors (high-speed photo-emitters) to optically encode the environment volume being tracked. Our research will be focused on the main program requirements by reducing cost, weight, footprint, and complexity of the radio enabled photo-sensing markers while increasing the overall resolution of the system (< 2 mm) so that we can use the positional information collected from multiple markers on an object to compute their orientation with a very high degree of accuracy (< 0.05 degree). The whole system would work within the IR spectrum and would therefore be invisible to the naked eye. For operation with NVG or IR vision augmentation systems we will need to further evaluate potential interference, and potentially adapt the design to support multiple modes of operation.
Benefit: The main advantages of our approach are: low manufacturing cost of the basic system components, extremely high accuracy (position: < 2 mm, orientation: < 0.05 degree), extremely high sampling frequency (> 120Hz), very low latency (< 10 ms), semi-automatic system calibration (easy setup), and an architecture that allows for easy upgrades and system expansion. All of these advantages point towards an excellent solution that can meet the most stringent requirements of AR applications to support infantry training now and in the future.
Keywords: Simulation, Augmented Reality, Low-cost, Training, motion tracking, Virtual Reality
