SBIR-STTR Award

The development of two different types of image sensors integrated in image processing computers is proposed. The first kind of sensors is an advanced 512x512 sized photodiode array with locally adaptive integration time control. The specialty of the local control is that it depends on the illumination of the neighboring pixels. The adaptive sensor control method and the neighborhood size are fully programmable. This sensor array will be sensitive in the visual range and also in the lower end of the near IR. The sensor array will be embedded in a CNN technology based focal plane sensor-processor array. This large size processor array will have close to 1 TerOPS computational power, and it will enable above 10,000 Fps operation. It will have a roughly 1000 times speed advantage compared to a 3GHz Pentium processor, while its power consumption will be only a few watts. The second kind of sensor will be an array of nanoantennas and nonlinear detectors for direct detection at wavelengths from the far infrared into the THz regime. The design of nanoantennas for long-wavelength infrared and THz radiation will be explored, and evaluation of possible nonlinear devices for direct detection at the antenna feed will be performed. Nano-mesa Si Schottky detectors and metal-oxide-metal tunnel diodes will be evaluated for use as the detector, with a down-selection to the most promising device candidate and antenna design for co-integration with the CNN technology based processor

This STTR Phase II proposal is based on work completed under the STTR Phase I grant entitled 'Focal Plane array processors with adaptive visual range and millimeter wave sensors' (Contract No. N00014-04-M-0212). In this Phase II project, we propose the development of two lines of ultra-high speed focal plane sensor and processor arrays equipped with (i) adaptive image sensors sensitive in the visual, NIR range, and (ii) uncooled nanoantenna MOM detector array sensitive in the far IR/THz domain respectively. The extraordinary computational performance of the digital CNN processor array will enable the capturing and processing of up to 10,000 frames per second in real-time in many image processing tasks. The sensor-processor chips will be integrated into a standalone vision system (demonstrator) and the capabilities of the new devices will be demonstrated via a high-speed multiple target tracking algorithm.

Benefit:
1. Development of a high resolution (well above 256x256), ultra-high speed (up to 10,000 FPS) CNN/CVM based focal plane processor array equipped with an adaptive optical input sensitive in the visual and NIR. This chip will have huge potential in real-time security and surveillance systems, visual multi-target tracking systems, and industrial quality control. 2. Development of uncooled nanoantennas, sensitive in the far infrared (THz) domain, and integrate them to an ultra-high speed, 8x8 and later 32x32 sized CNN/CVM chip. This technology will facilitate economical multi-spectral imaging systems usable (among others) in a number of security and surveillance applications. 3. Development of a compact (not larger than 7x4x12 inches), standalone vision system (demonstrator) that integrates both of the ultra-high speed sensors and demonstrates the capabilities of the system though various applications, including multi-spectral image fusion, multi-target tracking, and feature classification. This mobile and portable system will be able to demonstrate the capabilities of the adaptive CVM and the CVM_IR chip not only in a laboratory environment, but also in any indoor or outdoor situation.

Keywords:
millimeter wavelength imager, locally adaptive sensor array, focal plane processor array, nanoantenna, CNN/CVM, ultra-high speed vision system, high-speed image processing, Multi-Spectral Imaging