Stand-off facial identification capabilities are valuable for a number of scenarios, including security, border protection, and covert surveillance. Long-range cameras coupled with advanced facial recognition software can provide an automated means of finding and identifying people of interest.
Long-range imagery often suffers from atmospheric turbulence-based scintillation and warping. Other conditions, such as poor lighting and haze, can further degrade image quality to the point that too much information is lost to make a positive identification.
ATCOM includes techniques such as turbulence mitigation, local area enhancement, super resolution, and dehazing, all of which can be used to extract the additional information for imagery needed by facial recognition software to make a match. Through field testing with third party tools, ATCOM has improved the identification capabilities of leading facial recognition software.
Note: Tests performed with the ROC SDK from Rank One Computing.