Advanced Optical Systems, Inc. developed the ULTOR(r) system, a real-time correlation processor that looks for improvised explosive devices (IED) by examining imagery of vehicles. The system determines the level of threat an approaching vehicle may represent. The system works on incoming video collected at different wavelengths, including visible, infrared, and synthetic aperture radar. Sensors that attach to ULTOR can be located wherever necessary to improve the safety around a checkpoint. When a suspect vehicle is detected, ULTOR can track the vehicle, alert personnel, check for previous instances of the vehicle, and update other networked systems with the threat information.
The ULTOR processing engine focuses on the spatial frequency information available in the image. It correlates the imagery with templates that specify the criteria defining a suspect vehicle. It can perform full field correlations at a rate of 180 Hz or better. Additionally, the spatial frequency information is applied to a trained neural network to identify suspect vehicles.
We have performed various laboratory and field experiments to verify the performance of the ULTOR system in a counter IED environment. The experiments cover tracking specific targets in video clips to demonstrating real-time ULTOR system performance. The selected targets in the experiments include various automobiles in both visible and infrared video.
Advanced Optical Systems, Inc. has developed the ULTOR system, a compact, high-speed optical processor that performs object recognition and precision tracking in real time for real-world applications by using data from imaging sensors. The heart of the ULTOR target recognition and tracking system is an optical correlator. The system includes real-time preprocessing, large filter stores, filter management logic, correlation detection and thresholding, correlation tracking, and data output. It is self contained, receiving operational commands as an Internet appliance.
We have performed various laboratory and field experiments to verify the performance of the ULTOR system in a maritime environment. The experiments cover tracking specific objects in video clips to demonstrating real-time ULTOR system performance on the Cooper River in Charleston, South Carolina. The selected objects in the experiments include individual people, wave runners, boats, and ships. The paper will present a description of the ULTOR system and the results of the experiments. The latest processor advancements will be presented. These advancements will include ruggedization, algorithm development, and new operational methodologies.
The tragic loss of Space Shuttle Columbia threw the future of Hubble Space Telescope (HST) in doubt. The Columbia Accident Investigation Board report led NASA to the realization that astronauts must have someplace to go on orbit if the Shuttle is damaged, a requirement that cannot be met for a manned HST mission. Yet missions to HST are required, since HST was designed to be serviced periodically.
To address this problem, NASA is developing a robotic servicing mission to Hubble. On-orbit rendezvous and docking under tele-robotic or fully autonomous control involves a number of challenges that have not been fully resolved. One key challenge is how to bring two craft together in precise alignment to each other without an experienced astronaut on board. For this to be possible, sensors are needed to report relative distance, bearing, and orientation.
At Advanced Optical Systems (AOS), we have applied our ULTOR digital correlation system to the Hubble repair mission. The ULTOR system operates at approximately 10 Hertz and can accurately determine the relative distance, bearing, and orientation needed for semi- or fully-autonomous docking to HST. The system can operate using the HST berthing target or other features, including the HST itself. It is small and light enough to be placed on the servicing craft, thus avoiding orbit-to-ground communication latency issues. We will discuss the results of our testing with computer-generated imagery of the HST and with any hardware-in-the-loop simulations.