Object identification using optical code reading and object recognition

• US 9,424,480 B2
• Filed: 03/14/2013
• Issued: 08/23/2016
• Est. Priority Date: 04/20/2012
• Status: Active Grant

• Current Assignee: Datalogic Adc, Inc. (Datalogic S.p.A.)
• Original Assignee: Datalogic Adc, Inc. (Datalogic S.p.A.)
• Inventors: Hastings, Robert D.
• Primary Examiner(s): Mehta, Bhavesh
• Assistant Examiner(s): Mistry, Oneal R
• Application Number:
  US13/831,025
• Publication Number:
  US 20130279748A1
• Time in Patent Office: 1,258 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06V 10/768 using context analysis, e.g. recognition aided by known co-occurring patterns
• CPC Class Codes: G06V 2201/10 Recognition assisted with metadata

Abstract

An object identification system comprises an optical code reader that scans an optical code of an object and decodes a portion of the optical code. Using the decoded portion of the optical code, a database filter unit generates a filtered subset of feature models from a set of feature models of known objects stored in a database. An image capture device captures an image of the object, and a feature detector unit detects visual features in the image. A comparison unit compares the detected visual features to the filtered subset of feature models to identify a match between the object and a known object.

Claims

1. A method of identifying an object by use of a system that includes a database of feature models of known objects having known optical code information associated therewith, the feature models including visual features that are geometric point features corresponding to points on the known objects, the method comprising:

attempting to read an optical code on an object, the optical code encoding a set of characters;

decoding a portion of the optical code to obtain an item-filter parameter comprising incomplete information encoded in the optical code, the incomplete information corresponding to a decoded subset of the set of characters encoded in the optical code, the decoded subset including at least one member and fewer than all members of the set of characters;

when the item-filter parameter is obtained, generating a filtered subset of feature models from the database of feature models based on the item-filter parameter, each feature model of the filtered subset of feature models corresponding to a known object whose associated known optical code information includes the decoded subset;

capturing an image that includes a representation of the object, wherein the object is a physical item;

extracting visual features from the image, the extracted visual features being geometric point features corresponding to points on the object; and

comparing the extracted visual features to those of the feature models of the filtered subset of feature models to identify a match between the object and a known object.

2. The method of claim 1, further comprising:

detecting an exception event in response to the decoding yielding fewer than all of the members of the set of characters encoded in the optical code; and

generating the filtered subset of feature models in response to detecting the exception event.

3. The method of claim 1, wherein the decoded subset includes a check digit, and wherein each feature model of the filtered subset of feature models includes a known check digit equal to that of the decoded subset.

4. The method of claim 1, wherein the decoded subset includes a manufacturer identification, and wherein each feature model of the filtered subset of feature models includes a known manufacturer identification equal to that of the decoded subset.

5. The method of claim 1, wherein attempting to read the optical code on the object comprises reading a representation of the optical code from the image.

6. The method of claim 1, wherein the geometric point features are scale-invariant feature transform (SIFT) features.

7. The method of claim 1, wherein the geometric point features include multiple scale-invariant features defining a spatial relationship between the geometric point features.

8. The method of claim 1, wherein the decoded subset corresponds to information available in a fragment of the optical code.

9. The method of claim 1, wherein the object is an object of merchandise in a retail store.

10. An object identification system, comprising:

an optical code reader configured to decode a portion of an optical code on an object, and obtain from it an item-filter parameter comprising incomplete information encoded in the optical code, the incomplete information corresponding to a decoded subset of a set of characters encoded in the optical code, the decoded subset including at least one member and fewer than all members of the set of characters;

an image capture device configured to capture an image of the object, wherein the object is a physical item;

a feature detector unit configured to detect visual features in the image captured by the image capture device, the visual features detected in the image comprising geometric point features corresponding to points on the object;

a database configured to store feature models of known objects having known optical code information associated therewith, the feature models including visual features comprising geometric point features corresponding to points on the known objects;

a database filter unit in communication with the optical code reader and the database, the database filter unit configured to, when the item-filter parameter is obtained, generate a filtered subset of feature models based on the item-filter parameter; and

a comparison unit in communication with the feature detector unit and the database filter unit, the comparison unit configured to compare the detected visual features and the feature models of the filtered subset of feature models to identify a match between the object and a known object.

11. The object identification system of claim 10, wherein the feature detector unit, the database filter unit, and the comparison unit are implemented in a processor.

12. The object identification system of claim 10, wherein the optical code reader includes the image capture device.

13. The object identification system of claim 10, wherein the database is configured to store the known optical code information associated with the known objects.

14. The object identification system of claim 10, wherein the optical code reader is further configured to generate an exception event indicating that the optical code reader is unable to decode all of the optical code, and the database filter unit is further configured to generate the filtered subset of feature models in response to the exception event.

15. The object identification system of claim 10, wherein the database filter unit is configured to generate the filtered subset of feature models by identifying feature models in the database that correspond to known objects whose associated known optical code information includes the decoded subset.

16. The object identification system of claim 10, wherein the geometric point features are scale-invariant feature transform (SIFT) features.

17. The object identification system of claim 10, wherein the geometric point features include multiple scale-invariant features that define a spatial relationship between the geometric point features.

18. The object identification system of claim 10, wherein the decoded subset corresponds to information available in a fragment of the optical code.