# US 7,609,877 B2 · Tactical image parameter adjustment method for stereo pair correlation > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 7,609,877 B2 - Document ID: US7609877B2 - Status: Granted - Kind: B2 - Application number: 11/352,067 - Filed: 2006-02-09 - Published: 2007-08-09 - Granted: 2009-10-27 - Earliest claimed: 2006-02-09 - Assignees: The United States of America as represented by the Secretary of the Navy - Inventors: Benjamin Paul Carlson; William Rodney Ditzler; Michael R. Havlin - Web page: https://patentgrep.com/patent/US7609877B2 ## CPC - G01C 11/06: by comparison of two or more pictures of the same area - G06T 7/13: Edge detection - G06T 7/32: using correlation-based methods - G06V 10/255: {Detecting or recognising potential candidate objects based on visual cues, e.g. shapes} - G06V 20/13: Satellite images - G06T 2207/10012: Stereo images - G06T 2207/20056: Discrete and fast Fourier transform, [DFT, FFT] - G06T 2207/30181: Earth observation - G06T 2207/30212: Military - G01C: MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY (measuring liquid level G01F; radio navigation, determining distance or velocity by use of propagation effects, e.g. Doppler effects, propagation time, of radio waves, analogous arrangements using other waves G01S) - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract A method for the timely processing of tactical images which allows a military pilot adequate time to track and engage a target. The method uses a correlation process and back propagated neural network to home in on a correct position parameters for the tactical image including the heading and the range for the tactical image. ## Claims ### Claim 1 A tactical image parameter adjustment method for stereo pair correlation for determining a range for a tactical image comprising one or more computers implementing in the following steps: (a) providing tactical image position data generated by a sensor; (b) providing stereo image pairs which are stored within an image library wherein said stereo image pairs are based on a rough estimate of sensor position; (c) generating an edge image template from said tactical image position data for use in a correlation process; (d) generating first and second red and green combined edge templates in frequency spaces wherein each of said red and green combined edge templates consists of a preset number of differently weighted individual templates created using a parameter of interest; (e) correlating said edge image template with said first and second red and green combined edge templates to provide first and second correlation surfaces; (f) locating peak and minimum values in each of said frequency spaces in said first and second correlation surfaces; (g) subtracting the minimum value from the peak value for each of said frequency spaces in said first and second correlation surfaces; (h) generating first, second and third correlation values as inputs to a neural network wherein said first, second and third correlation are ratios of selected values of peak widths, peak heights and said peak values; (i) standardizing said first, second and third correlation values to fall between 0.0 and 1.0; and (j) providing said first, second and third correlation values standardized to fall between 0.0 and 1.0 to said neural network, wherein said neural network using said first, second and third correlation values determines a search direction, and then homes in on a correct value for the range of said tactical image. ### Claim 2 (depends on 1) The tactical image parameter adjustment method of claim 1 wherein each of said red and green combined edge templates consists of three individual weighted templates created using said parameter of interest±twenty percent of an operating range for said parameter of interest. ### Claim 3 (depends on 2) The tactical image parameter adjustment method of claim 2 wherein twenty percent of said operating range for said parameter of interest is initially set at 200 meters for the range of said tactical image. ### Claim 4 (depends on 1) The tactical image parameter adjustment method of claim 1 wherein said neural network uses a binary search to home in on said correct value for the range of said tactical image. ### Claim 5 (depends on 2) The tactical image parameter adjustment method of claim 2 wherein weighted values for the three weighted individual templates of said red edge template are 0.0, 0.5, and 1.0, and the weighted values for the three weighted individual templates of said green edge template are 1.0, 0.5 and 0.0. ### Claim 6 (depends on 1) The tactical image parameter adjustment method of claim 1 wherein said neural network comprises a back propagation neural network. ### Claim 7 (depends on 1) The tactical image parameter adjustment method of claim 1 wherein 2000 sets of said first, second and third correlation values are standardized for input to said neural network. ### Claim 8 (depends on 1) The tactical image parameter adjustment method of claim 1 wherein said first correlation values is a ratio of a red peak value over a green peak value, said second correlation value is a ratio of a red peak height over a red peak width, and said third correlation value is a ratio for a green peak height over a green peak width. ### Claim 9 A tactical image parameter adjustment method for stereo pair correlation for determining a range and a heading for a tactical image comprising one or more computers implementing in the following steps: (a) providing tactical image position data generated by a sensor; (b) providing stereo image pairs which are stored within an image library wherein said stereo image pairs are based on a rough estimate of sensor position; (c) generating an edge image template from said tactical image position data for use in a correlation process; (d) generating first and second red and green combined edge templates in frequency spaces wherein each of said red and green combined edge templates consists of a preset number of differently weighted individual templates created using a parameter of interest; (e) correlating said edge image template with said first and second red and green combined edge templates to provide first and second correlation surfaces; (f) locating peak and minimum values in each of said frequency spaces in said first and second correlation surfaces; (g) subtracting the minimum value from the peak value for each of said frequency spaces in said first and second correlation surfaces; (h) generating first, second and third correlation values as inputs to a pair of back-propagated neural network wherein said first, second and third correlation are ratios of selected values of peak widths, peak heights and said peak values; (i) standardizing said first, second and third correlation values to fall between 0.0 and 1.0; and (j) providing a first set of said first, second and third correlation values standardized to fall between 0.0 and 1.0 to said pair of back propagated neural networks, wherein said pair of back propagated neural network determines a search direction upon receiving said first set of said first, second and third correlation values; (k) providing a plurality of additional sets of said first, second and third correlation values to said pair of back propagated neural networks, each of said pair of back propagated neural networks using the additional sets of said first, second and third correlation values to indicate when a parameter value for said parameter of interest is larger or smaller than a correct position parameter for said parameter of interest; and (l) adjusting the parameter value for said parameter of interest until said correct position parameter for said parameter of interest is obtained by looping through steps (d)-(l) of said tactical image parameter adjustment method. ### Claim 10 (depends on 9) The tactical image parameter adjustment method of claim 9 wherein each of said red and green combined edge templates consists of three weighted individual templates created using said parameter of interest±twenty percent of an operating range for said parameter of interest. ### Claim 11 (depends on 10) The tactical image parameter adjustment method of claim 10 wherein twenty percent of said operating range for said parameter of interest is initially set at 200 meters when the range of said tactical image is said parameter of interest. ### Claim 12 (depends on 10) The tactical image parameter adjustment method of claim 10 wherein twenty percent of said operating range for said parameter of interest is initially set at one degree when the heading for said tactical image is said parameter of interest. ### Claim 13 (depends on 10) The tactical image parameter adjustment method of claim 10 wherein weighted values for the three weighted individual templates of said red edge template are 0.0, 0.5, and 1.0, and the weighted values for the three weighted individual templates of said green edge template are 1.0, 0.5 and 0.0. ### Claim 14 (depends on 9) The tactical image parameter adjustment method of claim 9 wherein 2000 sets of said first, second and third correlation values are standardized for input to said back propagated neural network. ### Claim 15 A tactical image parameter adjustment method for stereo pair correlation for determining a range and a heading for a tactical image comprising one or more computers implementing in the following steps: (a) providing tactical image position data generated by a sensor; (b) providing stereo image pairs which are stored within an image library wherein said stereo image pairs are based on a rough estimate of sensor position; (c) generating an edge image template from said tactical image position data for use in a correlation process; (d) generating first and second red and green combined edge templates in frequency spaces wherein each of said red and green combined edge templates consists of three weighted individual templates created using a parameter of interest, wherein weighted values for the three weighted individual templates of said red edge template are 0.0, 0.5, and 1.0, and the weighted values for the three weighted individual templates of said green edge template are 1.0, 0.5 and 0.0; (e) correlating said edge image template with said first and second red and green combined edge templates to provide first and second correlation surfaces; (f) locating peak and minimum values in each of said frequency spaces in said first and second correlation surfaces; (g) subtracting the minimum value from the peak value for each of said frequency spaces in said first and second correlation surfaces; (h) generating first, second and third correlation values as inputs to a pair of back-propagated neural network wherein said first, second and third correlation are ratios of selected values of peak widths, peak heights and said peak values; (i) standardizing said first, second and third correlation values to fall between 0.0 and 1.0; and (j) providing a first set of said first, second and third correlation values standardized to fall between 0.0 and 1.0 to said pair of back propagated neural networks, wherein said pair of back propagated neural network determines a search direction upon receiving said first set of said first, second and third correlation values; (k) providing a plurality of additional sets of said first, second and third correlation values to said pair of back propagated neural networks, each of said pair of back propagated neural networks using the additional sets of said first, second and third correlation values to indicate when a parameter value for said parameter of interest is larger or smaller than a correct position parameter for said parameter of interest; (l) adjusting the parameter value for said parameter of interest by adjusting a step size for said parameter of interest by approximately fifty percent until said correct position parameter is obtained, said correct position parameter being obtained by looping through steps (d)-(l) of said tactical image parameter adjustment method. ### Claim 16 (depends on 15) The tactical image parameter adjustment method of claim 15 wherein the first of said pair of back propagated neural networks uses a binary search to home in on the correct position parameter for the range of said tactical image and the second of said pair of back propagated neural networks uses said binary search to home in on the correct position parameter for the heading for said tactical image. ### Claim 17 (depends on 15) The tactical image parameter adjustment method of claim 15 wherein each of said red and green combined edge templates consists of the three individual weighted templates created using said parameter of interest±twenty percent of an operating range for said parameter of interest. ### Claim 18 (depends on 17) The tactical image parameter adjustment method of claim 17 wherein twenty percent of said operating range for said parameter of interest is initially set at 200 meters when the range of said tactical image is said parameter of interest. ### Claim 19 (depends on 17) The tactical image parameter adjustment method of claim 17 wherein twenty percent of said operating range for said parameter of interest is initially set at one degree when the heading for said tactical image is said parameter of interest. ### Claim 20 (depends on 15) The tactical image parameter adjustment method of claim 15 wherein 2000 sets of said first, second and third correlation values are standardized for input to said back propagated neural network.