Image processing of aerial imagery for energy infrastructure analysis

• US 10,467,473 B1
• Filed: 05/29/2019
• Issued: 11/05/2019
• Est. Priority Date: 01/14/2019
• Status: Active Grant

• Current Assignee: Ponderosa Advisors LLC
• Original Assignee: Sourcewater, Inc.
• Inventors: Adler, Joshua
• Primary Examiner(s): Tsai, Tsung Yin
• Application Number:
  US16/425,192
• Time in Patent Office: 160 Days
• CPC Class Codes: G06F 16/29 Geographical information databases
• CPC Class Codes: G06F 16/51 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/587 using geographical or spatial information, e.g. location
• CPC Class Codes: G06F 18/217 Validation; Performance evaluation; Active pattern learning techniques
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/10 using kernel methods, e.g. support vector machines [SVM]
• CPC Class Codes: G06N 20/20 Ensemble learning
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06V 10/776 Validation; Performance evaluation
• CPC Class Codes: G06V 20/176 Urban or other man-made structures

Abstract

A computer-implemented method for processing images to identify Energy Infrastructure (EI) features within aerial images of global terrain is provided. The image processing method identifies information about EI features by applying an EI feature recognition model to aerial images of global terrain. The EI feature recognition model identifies the EI feature information according to image content of the aerial image and according to supplemental information about EI features in the global terrain. The method further uses the results of the identification to update the EI feature recognition model.

Claims

1. A computer-implemented method for processing images to identify Energy Infrastructure (EI) features within aerial images of global terrain, the method to be carried out by at least one processor executing computer instructions, the method comprising:

receiving a working image comprising at least one aerial image of a first portion of global terrain;

applying an EI feature recognition model to the working image to identify feature-level information on a first EI feature of a first EI feature type at a first confidence level according to image content of the working image;

receiving, from a supplemental information source, information on a second EI feature of a second EI feature type, the second EI feature located within the first portion of global terrain;

updating the feature-level information on the first EI feature to a second confidence level based on the first confidence level and the information on the second EI feature;

modifying, during a training mode of operation, the EI feature recognition model, based on at least one training image of the portion of global terrain and on the information on the first EI feature at the second confidence level.

2. The computer-implemented method of claim 1 wherein the supplemental information source is an EI feature database and the information on the second EI feature is comprised within at least one EI feature data record stored within the EI feature database.

3. The computer-implemented method of claim 1 wherein the supplemental information source is a public records information source and the information on the second EI feature is comprised within at least one public record stored within the public records information source.

4. The computer-implemented method of claim 1 wherein the supplemental information source includes information on one of:

oilfield permits,

easements,

building permits,

energy infrastructure development proposals, consents or reports

maps,

deeds,

prospective land transactions,

land use proposals or consents,

mineral rights,

contracts,

news and media,

seismology information,

field data,

weather data,

legal information,

GPS and location information,

sensor data,

hydrological information.

5. The computer-implemented method of claim 1 wherein the first EI feature type or the second EI feature type include one or more of:

an EI development site,

a frac-water pit, frac pond or frac water impoundment,

a well pad,

a drilling rig,

pipeline infrastructure,

a service road,

a clearing,

a vehicle or truck,

a tank battery,

a proppant store,

a drilling reserve pit,

a frac spread,

a sand mine,

a producing well,

a flare system,

solar panel mounts,

solar panels,

an electrical substation,

a security fence,

a building,

a cable system,

a wind energy collector,

meteorological monitoring equipment,

construction equipment,

hydroelectric reservoirs or forebays,

hydroelectric intake structures,

penstocks,

surge chambers,

a hydroelectric power house,

a hydroelectric tailrace.

6. The computer-implemented method of claim 5 wherein the first EI feature type is the same as the second EI feature type.

7. The computer-implemented method of claim 5 wherein the first EI feature type is different to the second EI feature type.

8. The computer-implemented method of claim 1 wherein the information on the first EI feature or the information on the second EI feature comprises one or more of:

a presence or a suspected-presence of an EI feature,

a type of a known or suspected EI feature,

a status attribute of a known or suspected EI feature.

9. The computer-implemented method of claim 8 wherein the status attribute comprises one or more of:

a detection date or time,

a location,

a type,

an area, length, width, height, size or volume,

a number,

a number of sub-parts,

a surface rights owner,

a mineral rights owner,

a fluid level or volume,

a fluid or material type,

a fluid or material color,

a fluid or material quality,

a fluid or material attribute,

a surface type,

a level of activity, inactivity or operation,

a brightness, intensity or spectral content,

a connectivity,

a power delivery or capability,

a bore size,

a flow rate.

10. The computer-implemented method of claim 1 wherein the training image is the same as the working image.

11. The computer-implemented method of claim 1 wherein the training image is different to the working image.

12. The computer-implemented method of claim 1 wherein an image resolution of the training image is higher than an image resolution of the working image.

13. The computer-implemented method of claim 1 wherein modifying the EI feature recognition model includes training a neural network to establish neural network weights.

14. The computer implemented method of claim 1 wherein the information on the second EI feature is associated with a third confidence level, the method further comprising identifying the information on the first EI feature at the second confidence level based on the first confidence level, the information on the second EI feature and the third confidence level.

15. The computer-implemented method of claim 1 further comprising applying the modified EI feature recognition model to a further working image to identify a further EI feature of the first EI feature type.

16. The computer-implemented method of claim 1 wherein:

the training image includes an image time-sequence, the image time-sequence including a plurality of aerial images of the portion of global terrain, each aerial image of the plurality taken at one of a respective plurality of image capture times,

modifying the EI feature recognition model includes determining a pattern of stages of EI feature development across the image time-sequence.

17. The computer-implemented method of claim 1 wherein:

the working image includes an image time-sequence, the image time-sequence including a plurality of aerial images of the portion of global terrain, each aerial image of the plurality taken at one of a respective plurality of image capture times,

applying the EI feature recognition model includes recognizing from the image time-sequence, a stage of EI feature development across the image time-sequence.

18. A computer-implemented method for processing images to identify Energy Infrastructure (EI) features within aerial images of global terrain, the method to be carried out by at least one processor executing computer instructions, the method comprising:

receiving a training image comprising an image time-sequence, the image time-sequence including a plurality of aerial images of a portion of global terrain, each aerial image of the plurality taken at one of a respective plurality of image capture times;

receiving, from a supplemental information source, information on EI features located within the portion of global terrain;

generating an EI feature recognition model according to the image time-sequence of the training image and the information on EI features located within the portion of global terrain;

wherein the information on EI features located within the portion of global terrain identifies the appearance, presence-of or recording-of an EI feature at a historical time that is between the earliest and the latest of the plurality of image capture times.

19. The computer-implemented method of claim 18 further comprising applying the EI feature recognition model to a working image to identify an EI feature.

20. The computer-implemented method of claim 19 wherein:

generating the EI feature recognition model includes determining a pattern of stages of EI feature development across the image time-sequence, and

applying the EI feature recognition model includes recognizing a stage of EI feature development of the EI feature in the working image.