clarke@csri.toronto.edu (Jim Clarke) (03/23/88)
S P E C I A L J O I N T S E M I N A R sponsored by the Departments of Computer Science and Electrical Engineering Thursday, March 31, 10 a.m., WB119 (WB = Wallberg Building, 184 College Street) Dr. James F. Boulter National Defence "Digital Image Processing for Space-Based Infrared Surveillance" Canada is examining the feasibility of using a space-based infrared (IR) surveillance system in applications such as North American air defence, ensuring territorial sovereignty and for search and rescue. Because of their capability for detecting moving targets, the new generation of star- ing mosaic infrared (IR) detector arrays are prime candidates for use in such a future surveillance system. These sensors acquire a frame of data by simultaneously exposing all the detectors in a two- dimensional array to the scene radiance, and then sequentially reading them out. This yields image data with characteristics that are quite different from that acquired with traditional IR sensors which obtain imagery by mechanical or elec- tronic scanning. New digital image-processing techniques and concepts must be developed to calibrate and analyse the information obtained from this new class of imaging sensor. In addition, optimal strategies must be defined for scanning the sensor field-of-view over a large area and analys- ing the data in order to satisfy specific surveillance requirements. We begin with an overview of Canadian participation in a joint Allied Nation space-based IR surveillance program, and follow with an outline of the basic concepts and components of a functional system. Next, we describe R7D that the Defence Research Establishment Valcartier (DREV) is conducting in two particular areas of processing and simulation for space- based IR surveillance. We close with a brief summary of the image- processing facilities available at DREV for these studies. The first are to be described involves the development of computer models and system simulations to evaluate the performance tradeoffs and overall feasibility of selected surveillance configurations. It includes the development and validation (by using experimental measurements) of models that describe the IR characteristics of targets, backgrounds and the atmo- sphere. These models will be utilized to generate simulated data to develop and evaluate new processing algorithms, and will support a more general systems study of space-based IR surveillance to investigate candi- date satellite constellations, scanning strategies, sensor characteristics, signal processing algorithms, etc. The other R&D area that we will address involves the specific processing algorithms required for target enhancement and detection, automatic track- ing, target classification and identification, and multi-sensor data fusion. For each case we state the objectives of the processing, give an overview of processing techniques useful for achieving the desired results, show processing examples using simulated and experimental imagery, and identify productive areas for future research efforts. -- Jim Clarke -- Dept. of Computer Science, Univ. of Toronto, Canada M5S 1A4 (416) 978-4058 BITNET,CSNET: clarke@csri.toronto.edu CDNNET: clarke@csri.toronto.cdn UUCP: {allegra,cornell,decvax,linus,utzoo}!utcsri!clarke