TeraOps Corporation proposes to develop techniques for reproducible coloring of nighttime imagery and to implement a color night vision device based on TeraOps' powerful CNN array processing technology. In addition to CNN, the unit will combine other state-of-the-art technologies including miniature image intensifiers and high-contrast head mounted displays to create a low-power, portable unit that meets operational requirements. TeraOps will exploit the unique architecture of the CNN supercomputer-on-a-chip and its more than 10 to the 12th operations per second computing speed to implement the developed algorithmic techniques in real-time. These spatial-temporal image-processing algorithms will perform enhancement and adaptive, reproducible coloring of night imagery input through the integrated multi-band sensor. The algorithms will draw upon studies of biological image processing as well as on CNN models for color constancy and color visualization developed earlier by TeraOps researchers Colorized night vision goggles will greatly enhance the navigational and target recognition capabilities of military personnel. Color night vision units are expected to be of significant use for commercial aviation, police, rescue, and other civilian surveillance applications. Moreover, the CNN computer is a general- purpose machine that can be programmed to perform a variety of sophisticated image processing functions. The development of the CNN-based image processing engine will facilitate the development of products for real-time medical imaging (ultrasonic, MRI, CT, etc.), high-resolution displays, augmented reality systems, "smart" video cameras for automatic image acquisition, etc. The CNN computer will also be applicable to high-speed image processing for multimedia and video conferencing.
