A REAL TIME SYSTEM FOR MULTI-SENSOR IMAGE ANALYSIS THROUGH PYRAMIDAL SEGMENTATION

Award Information
Agency:
Department of Defense
Amount:
$50,000.00
Program:
SBIR
Contract:
N/A
Solitcitation Year:
N/A
Solicitation Number:
N/A
Branch:
Air Force
Award Year:
1991
Phase:
Phase I
Agency Tracking Number:
14890
Solicitation Topic Code:
N/A
Small Business Information
Cognitech Inc
2020 Broadway Suite 201, Santa Monica, CA, 90404
Hubzone Owned:
N
Woman Owned:
N
Socially and Economically Disadvantaged:
N
Duns:
N/A
Principal Investigator
 Leonid Rubin
 Principal Investigator
 (213) 453-6961
Business Contact
Phone: () -
Research Institution
N/A
Abstract
VOLUMETRIC INTERPRETATION OF MULTI-SPECTRAL SENSOR DATA AS WELL AS OTHER IMAGE ANALYSIS TASKS INVARIABLY INVOLVE A SEGMENTATION PREPROCESSING STEP. THE AIM OF THIS PROPOSAL IS TO IMPLEMENT A FAST (REAL-TIME) AND UNIVERSAL SEGMENTATION ALGORITHM FOR MULTI-SENSOR AND MULTISCALE IMAGE DATA. THE PROPERTIES OF THIS ALGORITHM ARE THE FOLLOWING. IT IS MULTISCALE AND PYRAMIDAL. IN OTHER WORDS, IT WILL NOT ONLY COMPUTE ONE SEGMENTATION, BUT A HIERARCHY OF SEGMENTATIONS FROM FINE TO COARSE SCALES. THE COARSER SEGMENTATION WILL BE DEDUCED FROM THE FINER ONE BY "MERGING" OPERATIONS, GIVING THUS A PYRAMIDAL STRUCTURE TO THE COMPUTATIONS. MOREOVER WITHIN EACH SCALE, VARIOUS SENSORS DATA IS INTEGRATED "FUSED". AS A CONSEQUENCE OF THIS PYRAMIDAL STRUCTURE, THE ALGORITHM WILL BE O(N) -A LINEAR TIME ALGORITHM (ON A SEQUENTIAL COMPUTER), THAT IS, THE COMPUTATIONAL TIME WILL BE EXACTLY PROPORTIONAL TO THE SIZE OF THE DATUM OR (LOG N) IN CONCURRENT IMPLEMENTATION. THE ALGORITHM IS UNIVERSAL, IT DOES NOT DEPEND ON ANY A PRIORI KNOWLEDGE ABOUT IMAGE STATISTICS. THE ALGORITHM IS WELL-POSED AND CONVERGES TO THE PROPER SOLUTION.

* information listed above is at the time of submission.

Agency Micro-sites

US Flag An Official Website of the United States Government