IP Library › Granted Patent US 11,861,843
Granted Patent B2
US 11,861,843 · App. 18/098,829 · Granted Jan 2, 2024

System and method for object analysis

Inventors: Matthieu Portail (Palo Alto, CA); Christopher Wegg (Palo Alto, CA); Mohammad-Ali Nikouei (Palo Alto, CA)
Assignee: Cape Analytics, Inc.
G06T7/11G01C11/06G06F16/432G06F16/489G06F18/22G06T7/0004G06T7/60G06T7/70G06T2207/10024G06T2207/20016G06T2207/20021G06T2207/20081G06T2207/20084G06T2207/30184G06T2207/30242
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 11,861,843
App. No.
18/098,829
Filed
Jan 19, 2023
Granted
Jan 2, 2024
Kind
B2
Art Unit
2669
USPC
382/276
Abstract

In variants, the method can include: determining a timeseries of measurements of a geographic region; determining a set of object representations from the timeseries of measurements; and determining a timeseries of object versions based on relationships between the object representations.

Claims (34)

1. A method comprising, by a processing system:

extracting a timeseries of building representations for a geographic region from a timeseries of measurements of the geographic region;

determining relationships between the building representations;

determining a timeseries of building versions based on the relationships between the building representations, wherein each building version is defined by a set of temporally adjacent and substantially similar building representations; and

determining, a timeseries of changes between the building versions based on the respective building representation sets, wherein the change between a prior and subsequent building version is labeled as:

modified when the first building version shares building segments with the second building version;

replaced when the first building version shares no building segments with the second building version;

added when a prior building representation set represented no building and a subsequent building representation represents a building; or

removed when the prior building representation represents a building and the subsequent building representation represents no building.

2. The method of claim 1 , further comprising associating, by the processing system, auxiliary information with the building versions based on a relationship between the auxiliary information and the respective building representations associated with the building version, wherein the auxiliary information is returned responsive to a request identifying the building version.

3. The method of claim 2 , wherein herein the auxiliary information associated with the building version is spatiotemporally related to the respective building representations.

4. The method of claim 2 , wherein the auxiliary information and the respective building representation are extracted from the same measurement set.

5. The method of claim 1 , wherein the changes between the building versions are inferred using a trained machine learning model.

6. The method of claim 1 , further comprising returning, by the processing system, an analysis of the changes between a first and second building version within the timeseries of building versions responsive to a request identifying the first building version.

7. The method of claim 1 , wherein the measurements comprise aerial images and the geographic region comprises a property parcel.

8. The method of claim 1 , wherein different building versions have different physical building geometries.

9. The method of claim 1 , wherein the building representations comprise building segments.

10. The method of claim 1 , wherein the building segments are components of the building.

11. The method of claim 1 , wherein the building segments comprise measurement segments.

12. The method of claim 1 , wherein the building representations comprise feature vectors extracted by a model trained to be agnostic to appearance-based changes and sensitive to structural changes.

13. A system, comprising a processing system configured to:

extract, by the processing system, building representations for each of a set of geographic regions from measurements of the respective geographic region;

identify, by the processing system, building versions based on sets of substantially similar and temporally adjacent building representations for a geographic region;

determine, by the processing system, spatiotemporally associated auxiliary information with each building version, wherein the spatiotemporally associated auxiliary information comprises a timeseries of changes between the building versions based on the respective building representation sets, wherein the change between a prior and subsequent building version is labeled as:

modified when the first building version shares building segments with the second building version;

replaced when the first building version shares no building segments with the second building version;

added when a prior building representation set represented no building and a subsequent building representation represents a building; or

removed when the prior building representation represents a building and the subsequent building representation represents no building;

store, by the processing system, the spatiotemporally associated auxiliary information with each building version; and

return, by the processing system, auxiliary information associated with a building version responsive to a request identifying the building version.

14. The system of claim 13 , wherein different building versions have different physical geometries.

15. The system of claim 13 , wherein the building representations comprise feature vectors extracted by a model trained to be agnostic to appearance changes and sensitive to geometric changes.

16. The system of claim 13 , wherein the system is further configured to determine, by the processing system, a set of relationships between the building versions based on the respective building representation sets, wherein an analysis based on the set of relationships is returned responsive to the request.

17. The system of claim 13 , wherein the building representations comprise building segments, wherein the relationships are determined using a set of heuristics applied to a comparison between the building segments.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 23, 2023
From: PORTAIL, MATTHIEU; WEGG, CHRISTOPHER; NIKOUEI MAHANI, MOHAMMAD-ALI
To: CAPE ANALYTICS, INC.
Reel/Frame 062785/0360 →
Continuity (2)
Provisional Application 63300736 · Jan 19, 2022
Related Publication 20230230220A1 · Jul 20, 2023
Cited By (5)
US 12,229,845 US 12,243,301 US 12,293,579 US 12,518,395 US 12,743,844