US 11,380,080 B2Granted

Object ingestion through canonical shapes, systems and methods

AssigneeNant Holdings IP, LLC

InventorsKamil Wnuk, David McKinnon, Jeremi Sudol, Bing Song, Matheen Siddiqui

Life of this patent

Estimating the term…

Loading patent…

Abstract

An object recognition ingestion system is presented. The object ingestion system captures image data of objects, possibly in an uncontrolled setting. The image data is analyzed to determine if one or more a priori know canonical shape objects match the object represented in the image data. The canonical shape object also includes one or more reference PoVs indicating perspectives from which to analyze objects having the corresponding shape. An object ingestion engine combines the canonical shape object along with the image data to create a model of the object. The engine generates a desirable set of model PoVs from the reference PoVs, and then generates recognition descriptors from each of the model PoVs. The descriptors, image data, model PoVs, or other contextually relevant information are combined into key frame bundles having sufficient information to allow other computing devices to recognize the object at a later time.

Claims23

Dependent claims

An object ingestion system, comprising:

at least one non-transitory computer readable memory storing executable object ingestion software instructions;
and
at least one processor coupled with the at least one non-transitory computer readable memory that, upon execution of the object ingestion software instructions, performs operations to:
obtain a digital representation of a scene, wherein the digital representation is obtained from at least one sensor and further includes image data of at least one object and location information;
obtain a result set of shape objects from a canonical shape database, wherein the result set includes at least one shape object stored in the canonical shape database and that has at least one shape attribute satisfying selection criteria determined from geometrical information of the at least one object derived from the image data of the at least one object;
select at least one target shape object from the result set of shape objects based on a context associated with the at least one object;
instantiate at least one object model of the at least one object from the at least one target shape object and the image data;
and
store, in an object recognition database, a bundle of recognition parameters derived from the object model and location information, wherein the recognition parameters enable a computing device to recognize the at least one object.

6 more elements

Dependencies are inferred from the claim text.

Facts

Application: 17/040,000Filed 30 Sep 2020 · continuation
Priority: 14 Feb 2014
Published: US 2021/0027084 A128 Jan 2021
Granted: 5 Jul 2022
Term ends: Not estimatedThe source records don’t support an estimate.

Classification13 codes

G06V 20/46
In video content: extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
G06F 16/532
Querying: query formulation, e.g. graphical querying
G06F 16/5838
Using metadata automatically derived from the content: using colour
G06F 16/5854
Using metadata automatically derived from the content: using shape and object relationship

Family

Application61/940,320Us provisional application · filed 14 Feb 2014
Application14/623,435Parent of this continuation · filed 16 Feb 2015
Application15/297,053Parent of this continuation · filed 18 Oct 2016
US 9,501,498 B2Parent of this continuation · filed 22 Nov 2016
Application16/123,764Parent of this continuation · filed 6 Sep 2018
US 10,095,945 B2Parent of this continuation · filed 9 Oct 2018
US 2021/0027084 A1Pre-grant publication · published 28 Jan 2021
US 10,832,075 B2Parent of this continuation
US 2015/0302027 A1Published application
US 2017/0039442 A1Published application
US 2019/0012557 A1Published application
US 11,380,080 B2This patent

From the US filing history. Foreign family members aren’t listed.