Introduction

The ULTOR® Object Recognition System is based on R&D investments made by AOS, the Navy, the Army, and NASA from 1988 through today. It has been demonstrated at sea, on UAV platforms, in ground applications, A variant of the ULTOR® system (the ULTOR® P3E) is currently manifested for a space mission to the Hubble Space Telescope in 2008.
 

Concept

ULTOR® ORS employs Fourier-class algorithms combined with both pre- and post-processing techniques to achieve excellent results from real-world imagery. Pre-processing algorithms such as edge-enhancement and local contrast enhancement are used to affect image properties. These image properties in turn affect the frequency spectrum in the Fourier transformed image.

Multiple Fourier domain algorithms may be used. These algorithms include filtering (including correlation) and power spectrum analysis. Various post-processing algorithms are used, depending upon which algorithms are chosen. For example, neural networks may be applied to power spectrum analysis, or correlation peak detection may be applied to maximize the signal-to-noise ratio of the correlation peak.

The object identification results are used to annotate a live video stream or to index a video database.
 

Implementation

ULTOR® ORS is an automatic target recognition algorithm set that has been ported to five different computing platforms. It has been implemented as software for a Windows platform, software on a real-time VxWorks platform, and a set of Field Programmable Array (FPGA) firmware.
 

Applications

ULTOR® technology has been matured to TRL 6-7 and has been tested in a variety of environments, including a space mission scheduled for late 2008. The images below show frames from a live tracking demonstration at Charleston Harbor, SC.
 

ULTOR® ORS recognizes and tracks a potential surface threat in a view from radar-cued infrared system.

For more information contact:

Joel Burcham
This e-mail address is being protected from spambots. You need JavaScript enabled to view it