Image analysis – Learning systems
Reexamination Certificate
1999-05-24
2002-07-23
Boudreau, Leo (Department: 2621)
Image analysis
Learning systems
C382S103000, C382S106000, C382S159000, C382S228000, C701S027000, C706S001000, C706S009000, C706S045000, C706S905000
Reexamination Certificate
active
06424736
ABSTRACT:
BACKGROUND
Extensive research has been performed on the visibility of objects in visual scenes and in infrared scenes of interest wherein the intent is to determine the probability that the object will be detected or to determine the amount of search time needed to do so. A 1995 doctoral thesis by one of the inventors herein discusses much of the research work known at that time. This thesis is “Modem Approaches to Computation of the Probability of Target Detection in Cluttered Environments” by Thomas J. Meitzler. The thesis proposed that a fuzzy logic approach could be employed in a method of calculating the probability of detecting targets using various field parameters and target metrics that included clutter metrics as input variables. Subsequent to the thesis, a patent application was filed which has resulted in a continuation application now before the US Patent Office. The continuation application is entitled “Method of Determining Probability of Target Detection in a Visually Cluttered Scene,” has Ser. No. 09/006,529, was filed on Jan. 13, 1998, and has two of the same inventors as the instant application. The parent application has the same title, lists Thomas J. Meitzler as the sole inventor, is now abandoned, had Ser. No. 08/614,087 and was filed Mar. 13, 1996. These prior applications related to predicting target detection but did not relate to the use of fuzzy logic approaches in doing so.
Meanwhile the inventors continued their work in fuzzy logic approaches to the study of target detection. As this work began, there was no certainty that a fuzzy logic approach could successfully be reduced to practice, but by September or October of 1998, a fuzzy logic approach had been sufficiently refined to be practical. Even further refinement was subsequently accomplished by the addition of a new input variable, which is the number of wavelet edge points in the scenes. As the work evolved papers were published which were co-authored by one or more of the inventors. These papers related to the use of fuzzy logic to predict either the probability of target detection or the amount of search time required to detect a target in a visual or infrared scene. These papers included the following publications.
1. “Fuzzy Logic Approach for Computing the Probability of Target Detection in Cluttered Scenes” by Thomas J. Meitzler, Labib Arefeh, Harpreet Singh and Grant R. Gerhart;
Optical Engineering.
35(12) 3623-3636 (December 1996).
2. “Predicting the Probability of Target Detection in Static Infrared and Visual Scenes using the Fuzzy Logic Approach” by Thomas J. Meitzler, Harpreet Singh, Labib Arefeh, Euijung Sohn and Grant R. Gerhart:
Optical Engineering.
37(1) 10-17 (January 1998).
3. “A Perception Test of Infrared Images of Soldiers in Camouflaged Uniforms” by Thomas J. Meitzler, Darin Ellis, Euijung Sohn, Darryl Bryk, Kathy Rock and Harpreet Singh; Proceedings of the SPIE's 12th Annual International Symposium on Aerospace Defense Sensing Simulation and Controls, Targets and Backgrounds Characterization & Representation IV, Vol. 3375, April 1998.
4. “Computing the Probability of Target Detection in Dynamic Visual Scenes Containing Clutter Using the Fuzzy Logic Approach” by Thomas J. Meitzler, Euijung Sohn, Daryl Bryk, R. Kistner and D. Bednarz; Optical Engineering, Vol. 37(7) 1998.
SUMMARY
Our invention is a method of determining the visibility of a target in a visual or infrared background scene where the target's visibility is a key design issue. Search time is the output variable used to gauge target detectability, and search time is an indicator of the probability of detecting the target in the background. We use a set of images of the target in the background scene and measure selected input variables in each image. The input variables relate to target size, target juxtaposition relative to the viewer's location, luminance of the target, and luminance of the background scene. Each input variable, as well as the output variable, is associated with its own set of empirically developed membership functions. These membership functions assign membership values to the variables. The choice of membership function for each variable is controlled by specially tailored fuzzy rules which always apply. Applying the fuzzy rules and membership functions results in a collection of values for the output variable which are subjected to a defuzzification, typically by centroid averaging. Defuzzification obtains a crisp value usable, among other things, to disqualify proposed target designs or to select among several qualified target designs.
REFERENCES:
patent: 5671140 (1997-09-01), Bessacini et al.
patent: 5867386 (1999-02-01), Hoffberg et al.
patent: RE36280 (1999-08-01), Kawabata et al.
Meitzler Thomas J.
Singh Harpreet
Sohn Euijung
Boudreau Leo
Carter Clifford C.
Chawan Sheela
Kuhn David L.
Soderling Gail S.
LandOfFree
Fuzzy logic technique to determine search time and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Fuzzy logic technique to determine search time and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Fuzzy logic technique to determine search time and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2819912