US 10,769,488 B1Granted

Item variation management

AssigneeAmazon Technologies, Inc.

InventorsSudarshan Narasimha Raghavan, Xiaofeng Ren, Michel Leonard Goldstein, Ohil K. Manyam

Life of this patent

Estimating the term…

Loading patent…

Abstract

This disclosure describes a system for automatically identifying an item from among a variation of items of a same type. For example, an image may be processed and resulting item image information compared with stored item image information to determine a type of item represented in the image. If the matching stored item image information is part of a cluster, the item image information may then be compared with distinctive features associated with stored item image information of the cluster to determine the variation of the item represented in the received image.

Claims20

Dependent claims

A computing system, comprising:

an item images data store;
one or more processors;
and
a memory coupled to the one or more processors and storing program instructions that when executed by the one or more processors cause the one or more processors to at least:
receive an image of an item as the item is received into a materials handling facility;
process the image of the item to generate first item image information that includes an arrangement of a first plurality of features;
determine a correlation score between the first item image information and a second item image information, wherein the second item image information is accessible from an item images data store;
determine that the correlation score exceeds a threshold;
determine distinctive features of the first item image information that are different than features of the second item image information;
associate at least one of the distinctive features or the first item image information with the second item image information to form a cluster corresponding to an item type of the item;
and
store the association, first item image information and the distinctive features in the item images data store.

9 more elements

A method to maintain a data store of item variations of an item, the method comprising:

receiving a first image of a first item;
processing the first image to determine a first arrangement of features of the first item;
comparing the first arrangement of features with a stored arrangement of features associated with a stored item image information to determine a correlation score indicative of a similarity between the first arrangement of features and the stored arrangement of features;
determining that the correlation score does not exceed a threshold;
in response to determining that the correlation score does not exceed the threshold:
storing first item image information in the data store as representative of the first item, wherein the first item image information includes the first arrangement of features;
subsequent to storing the first item image information:
receiving a second image of a second item;
processing the second image to determine a second arrangement of features of the second item;
comparing the second arrangement of features with the first item image information to determine a second correlation score indicative of a similarity between the second arrangement of features and the first arrangement of features;
determining that the second correlation score exceeds the threshold;
determining distinctive features of the second item image information that are different than features of the first item image information;
associating at least one of the distinctive features or the second item image information with the first item image information to form a cluster corresponding to an item type of the first item and the second item;
and
storing the cluster, including the association, the first item image information and the second item image information in the data store.

12 more elements

Dependencies are inferred from the claim text.

Facts

Application: 16/520,861Filed 24 Jul 2019 · continuation
Priority: 19 Sep 2013
Granted: 8 Sep 2020
Term ends: Not estimatedThe source records don’t support an estimate.

Classification3 codes

G06V 20/52
Context or environment of the image: surveillance or monitoring of activities, e.g. for recognising suspicious objects
G06V 10/751
Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g.…:…
G06V
Image and video recognition

Family

Application14/031,798Parent of this continuation · filed 19 Sep 2013
US 10,366,306 B1Parent of this continuation
US 10,769,488 B1This patent

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

Timeline

Citations

Open elsewhere

Building an agent or pulling many patents? Use the API or MCP server instead of scraping pages. Get an API key