SBIR-STTR Award

Determine the feasibility and system methodology of using helical scan in stereoscopic robot vision to replicate human vision with high a[ acuity in the foveal centralis. The resulting video from each robotic camera "eye" has the highest pick-up tube resolution possible at the center and a falling off in acuity to the periphery of the field-of-view. A two-color field sequential system is proposed. The video is digital and placed in a frame buffer, superimposed on the helical format of computer generated imagery (cgi). The difference image is processed to drive the robot. The cgi is the scenario of the task the robot is to perform, and the system keeps the robot "oncourse" with the task. The data base model for the cgi can be altered on-line from operator or output of artificial intelligence if the task is to deviate from a pre-ordained or "canned" program. The synergy of helical scan and cgi is a technical advantage by concentrating the perceived information in the robot pick-up raster and "* also concentrating the image processing of the computer in the "busy" areas of the scene. A high efficiency is gained in an intelligent machine by this mimicry of human visual sensing.

This project will design, construct and test the prototype development of helical scan in the imager and in the display. The composite vision system will comprise the parameters that were derived from phase i; namely a double spiral interlaced 2:1 with a clockwise and counter-clockwise field, each of 250 rings. The format will present an image update of 30 frames a second and a display refresh of 60 fields a second. With a lead time in the outside vendor development of solid state devices for imaging (ccd) and display (lcd), a laboratory model will be built for test and evaluation of stereopsis. The application will follow the requirements of a teleoperated vehicle for explosive ordnance disposal. Experiments in frame sequential color will be conducted on the monochromatic system.