IP Library Granted Patent US 11,417,087
Granted Patent B2
US 11,417,087 · App. 16/514,901 · Granted Aug 16, 2022

Image processing system including iteratively biased training model probability distribution function and related methods

Inventors: Michael P. Deskevich (Boulder, CO); Robert A. Simon (Longmont, CO); Christopher R. Lees (Northglenn, CO)
Assignee: HARRIS GEOSPATIAL SOLUTIONS, INC.
G06V20/13G06F17/18G06K9/6256G06K9/6267G06N20/00
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 11,417,087
App. No.
16/514,901
Granted
Aug 16, 2022
Kind
B2
Abstract

An image processing system may include a processor and an associated memory configured to store training data that includes training geospatial images. Each training geospatial image may include pixels. The processor may be configured to operate a training model to identify a given feature from each of the training geospatial images, and to iteratively generate a probability distribution function based upon a number of pixels corresponding to the given feature and also based upon a bias factor being reduced with each iteration.

Claims (32)

1. An image processing system comprising:

a processor and an associated memory configured to

store training data comprising a plurality of training geospatial images, each training geospatial image comprising a plurality of pixels, and

operate a training model to identify a given feature from each of the plurality of training geospatial images, and to iteratively generate a probability distribution function based upon a number of pixels corresponding to the given feature and also based upon a bias factor being reduced with each iteration.

2. The image processing system of claim 1 wherein the training model comprises a mask-based training model.

3. The image processing system of claim 1 wherein said processor is configured to reduce the bias factor to a terminal non-zero bias factor.

4. The image processing system of claim 1 wherein said processor is configured to reduce the bias factor to a terminal zero bias factor.

5. The image processing system of claim 1 wherein said processor is configured to use an initial bias factor so that the training model identifies a pixel corresponding to the given feature for each predetermined number of pixels that do not correspond to the given feature.

6. The image processing system of claim 1 wherein said processor is configured to generate the probability distribution function based upon an entropy of each training geospatial image.

7. The image processing system of claim 1 wherein said processor is configured to generate the probability distribution function based upon a Shannon entropy of each training geospatial image.

8. The image processing system of claim 1 wherein said processor is configured to operate the training model to identify the given feature from a plurality of geospatial images.

9. A method of processing an image comprising:

using a processor and an associated memory to

store training data comprising a plurality of training geospatial images, each training geospatial image comprising a plurality of pixels, and

operate a training model to identify a given feature from each of the plurality of training geospatial images, and to iteratively generate a probability distribution function based upon a number of pixels corresponding to the given feature and also based upon a bias factor being reduced with each iteration.

10. The method of claim 9 wherein the training model comprises a mask-based training model.

11. The method of claim 9 wherein using the processor comprises using the processor to reduce the bias factor to a terminal non-zero bias factor.

12. The method of claim 9 wherein using the processor comprises using the processor to reduce the bias factor to a terminal zero bias factor.

13. The method of claim 9 wherein using the processor comprises using the processor to use an initial bias factor so that the training model identifies a pixel corresponding to the given feature for each predetermined number of pixels that do not correspond to the given feature.

14. The method of claim 9 wherein using the processor comprises using the processor to generate the probability distribution function based upon an entropy of each training geospatial image.

15. The method of claim 9 wherein using the processor comprises using the processor to generate the probability distribution function based upon a Shannon entropy of each training geospatial image.

16. The method of claim 9 wherein using the processor comprises using the processor to operate the training model to identify the given feature from a plurality of geospatial images.

17. A non-transitory computer readable medium for processing an image, the non-transitory computer readable medium comprising computer executable instructions that when executed by a processor cause the processor to perform operations comprising:

storing training data comprising a plurality of training geospatial images, each training geospatial image comprising a plurality of pixels; and

operating a training model to identify a given feature from each of the plurality of training geospatial images, and to iteratively generate a probability distribution function based upon a number of pixels corresponding to the given feature and also based upon a bias factor being reduced with each iteration.

18. The non-transitory computer readable medium of claim 17 wherein the training model comprises a mask-based training model.

19. The non-transitory computer readable medium of claim 17 wherein the operations comprise reducing the bias factor to a terminal non-zero bias factor.

20. The non-transitory computer readable medium of claim 17 wherein the operations comprise reducing the bias factor to a terminal zero bias factor.

21. The non-transitory computer readable medium of claim 17 wherein the operations comprise using an initial bias factor so that the training model identifies a pixel corresponding to the given feature for each predetermined number of pixels that do not correspond to the given feature.

22. The non-transitory computer readable medium of claim 17 wherein the operations comprise generating the probability distribution function based upon an entropy of each training geospatial image.

23. The non-transitory computer readable medium of claim 17 wherein the operations comprise generating the probability distribution function based upon a Shannon entropy of each training geospatial image.

24. The non-transitory computer readable medium of claim 17 wherein the operations comprise operating the training model to identify the given feature from a plurality of geospatial images.

Assignments (4)
JOINDER TO PATENT SECURITY AGREEMENT Recorded Nov 5, 2025
From: NV5 GEOSPATIAL SOLUTIONS, INC.
To: JEFFERIES FINANCE LLC, AS COLLATERAL AGENT
Reel/Frame 073452/0582 →
CHANGE OF NAME Recorded Sep 25, 2023
From: L3HARRIS GEOSPATIAL SOLUTIONS, INC.
To: NV5 GEOSPATIAL SOLUTIONS, INC.
Reel/Frame 065218/0450 →
CHANGE OF NAME Recorded Jun 13, 2022
From: HARRIS GEOSPATIAL SOLUTIONS, INC.
To: L3HARRIS GEOSPATIAL SOLUTIONS, INC.
Reel/Frame 060346/0294 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 23, 2019
From: DESKEVICH, MICHAEL P.; LEES, CHRISTOPHER R.; SIMON, ROBERT A.
To: HARRIS GEOSPATIAL SOLUTIONS, INC.
Reel/Frame 049828/0775 →