SBIR-STTR Award

The ability to resolve closely spaced objects, which are separated by angles less than the diffraction limit of the observing platform, greatly facilitates the early discrimination of missile payloads. We demonstrate that our image processing provides accurate source positions and radiometry for sources separated by 0.25–0.5 of the full-width at half maximum (FWHM) of the point-spread function, or a factor of 2–4 closer than a separation of 1 FWHM or more, conventionally thought to be required for accurate detection. Furthermore, we show that this performance is at or near the theoretical limit derived from the signal-to-noise ratio of the data. We propose to design custom image analysis hardware that implements our image processing in a small fraction of a second. This hardware is expected to be relatively compact (a few printed-circuit boards) and require low power, allowing its use aboard spacecrafts. Future ASIC implementation will further reduce size and power. In addition to missile defense, the proposed hardware will find applications in a number of military and commercial markets, such as military intelligence, medical imaging, commercial satellite imaging, and high-end microscopy. Anticipated Benefits/Commercial Applications: Missile Defense, Military Intelligence, Surveillance, Law Enforcement, Commercial Satellite Imaging, Medical Tomography, Rational Drug Design, Microscopy.

Keywords:
Sub-Diffraction Imaging, Pixon Method, Image Reconstruction , Super-Resolution, Image Enhancement, Image Restoration

The Pixon method enables reliable and accurate imaging well below the diffraction limit. It therefore offers to missile defense two wholly new options for detection and tracking of closely-spaced objects (CSOs): (1) Pixon processing can increase resolution by a factor of four beyond the diffraction limit. It thus can resolve targets at ranges four times greater than otherwise possible, increasing the time available for critical decision making. (2) Conversely, telescope apertures can be reduced by a factor of four while holding performance constant, resulting in significantly reduced system cost. We propose to build a modular Pixon processor that will not only demonstrate such resolution enhancement but also significantly decrease noise and achieve theoretically optimum radiometry. The prototype-implemented in field-programmable gate arrays on a single circuit board-will process images of CSO clusters in 0.03 second. Multiple processors can work in parallel to process larger or multiple images; future implementations in application-specific integrated circuits will further increase speed. In an Enhancement option, we propose to interface this core processor with Aeromet, Inc.'s AIRS sensors and field test the full system. The hardware suits many additional applications: conventional surveillance (satellites or UAVs), guided munitions, satellite mapping, microscopy, and medical tomography