Polaris Sensor Technologies, Inc. is working on identifying single pixels in IR imagery at signal to noise (SNR) ranges from 1.25 to 3 with filters and neural networks. Some of these algorithms yield as high as 95% probability of detection in this low SNR range. Techniques that make use of matched filters, neural networks, and genetic algorithms are able to address this problem from a new point of view that may allow for development of a state of the art systems that operate with improved sensitivity. The methodologies developed by Polaris Sensor Technologies in this work can be applied to a variety of markets including computer analysis of medical imagery, computer vision for quality control, perimeter security, and reconnaissance

Polaris Sensor Technologies also has extensive work in image processing, and has developed and implemented techniques to perform:
    • Distortion removal in video imagery
    • Image registration among multiple cameras or video channels
    • Real time point ranging using camera video
    • Real time 3D image calculation for viewing on stereo LCD display