System and method for detecting an address block and barcode on a captured image of item, and reading the detected barcode using connected component analysis

Patent metadata

Field	Value
Publication number	US-12136017-B2
Application number	US-202318308529-A
Country	US
Kind code	B2
Filing date	Apr 27, 2023
Priority date	Apr 28, 2022
Publication date	Nov 5, 2024
Grant date	Nov 5, 2024

Abstract

The present disclosure relates to a system and method for detecting an address block and barcode on a captured image of an item, and reading the detected barcode using connected component analysis. In one aspect, the method includes binarizing a captured image to generate a binarized image having pixel values, inverting the pixel values of the binarized image, processing the inverted pixel values, and filtering the processed image by area. The method may also include machine learning the processed image to cluster objects in the processed image into a plurality of groups using the filtered image, and determining a number of objects in each of the clustered groups by performing additional operations on the clustered objects to detect the address block. The method may also include selecting the group having the greatest number of objects as the address block, and extracting the selected address block.

First claim

What is claimed is:

1. A method of detecting an address block from a captured image of an item, the method comprising:
receiving or retrieving, at a processor, a captured image of the item containing an address block;
inverting, at the processor, pixel values of the captured image;
processing, at the processor, the inverted pixel values of the captured image;
filtering, at the processor, the processed image by area;
machine learning, at the processor, the processed image to cluster objects in the processed image into a plurality of groups using the filtered image;
determining, at the processor, a number of objects in each of the clustered groups by performing additional operations on the clustered objects to detect the address block; and
selecting, at the processor, at least one of the clustered groups as the address block based on the determined number of objects.

2. The method of claim 1 , wherein the selecting comprises selecting a group of the at least one of the clustered groups having the greatest number of objects as the address block.

3. The method of claim 1 , wherein the selecting comprises selecting a group of the at least one of the clustered groups having a predetermined number of objects as the address block.

4. The method of claim 1 , wherein the selecting comprises:
selecting two or more groups in a predetermined number range of objects from the clustered groups; and
selecting a group having the greatest number of objects in the selected two or more groups as the address block.

5. The method of claim 1 , further comprising:
extracting, at the processor, the selected address block; and
storing, in a memory, the extracted address block.

6. The method of claim 5 , further comprising reading the extracted address block using an optical character recognition (OCR) process or artificial intelligence (AI).

7. The method of claim 1 , wherein the machine learning comprises performing unsupervised machine learning on the processed image to cluster the objects into the plurality of groups.

8. The method of claim 7 , wherein the unsupervised machine learning comprises a density-based spatial clustering of applications with a noise (DBSCAN) algorithm.

9. The method of claim 1 , wherein the filtering comprises filtering components of the processed image to a given range of sizes such that all components outside of minimum and maximum ranges are removed from the processed image.

10. The method of claim 1 , wherein the machine learning comprises identifying a barcode based on at least one of a width, a height, or an orientation of the objects being clustered.

11. A system for detecting an address block from a captured image of an item, the system comprising:
a memory storing instructions; and
one or more processors configured to execute the instructions to:
receive or retrieve a captured image of the item containing an address block;
invert pixel values of the captured image;
process the inverted pixel values;
filter the processed image by area;
perform machine learning on the processed image to cluster objects in the processed image into a plurality of groups using the filtered image;
determine a number of objects in each of the clustered groups by performing additional operations on the clustered objects to detect the address block; and
select at least one of the clustered groups as the address block based on the determined number of objects.

12. The system of claim 11 , wherein at least one of the one or more processors is configured to select a group of the at least one of the clustered groups having the greatest number of objects as the address block.

13. The system of claim 11 , wherein at least one of the one or more processors is configured to select a group of the at least one of the clustered groups having a predetermined number of objects as the address block.

14. The system of claim 11 , wherein at least one of the one or more processors is configured to:
select two or more groups in a predetermined number range of objects from the clustered groups; and
select a group having the greatest number of objects in the selected two or more groups as the address block.

15. The system of claim 11 , wherein at least one of the one or more processors is further configured to:
extract the selected address block; and
store, in a memory, the extracted address block.

16. The system of claim 11 , wherein at least one of the one or more processors is configured to perform unsupervised machine learning on the processed image to cluster the objects into the plurality of groups.

17. The system of claim 16 , wherein the unsupervised machine learning comprises a density-based spatial clustering of applications with a noise (DBSCAN) algorithm.

18. The system of claim 11 , wherein at least one of the one or more processors is configured to filter components of the processed image to a given range of sizes such that all components outside of minimum and maximum ranges are removed from the processed image.

19. The system of claim 11 , wherein at least one of the one or more processors is configured to identify a barcode based on at least one of a width, a height, or an orientation of the objects being clustered.

20. A non-transitory computer readable medium for storing instructions that cause, when executed by one or more processors, the one or more processors to:
receive a captured image of the item containing an address block;
invert pixel values of the captured image;
process the inverted pixel values of the captured image;
filter the processed image by area;
perform machine learning the processed image to cluster objects in the processed image into a plurality of groups using the filtered image;
determine a number of objects in each of the clustered groups by performing additional operations on the clustered objects to detect the address block; and
select at least one of the clustered groups as the address block based on the determined number of objects.

Assignees

• United States Postal Service

Inventors

• Simpson Ryan J

Classifications

• G06T2207/30242

  Counting objects in image · CPC title

• G06T7/60

  Analysis of geometric attributes · CPC title

• G06T5/20

  using local operators · CPC title

• G06K7/1413

  1D bar codes · CPC title

• G06V10/762

  using clustering, e.g. of similar faces in social networks · CPC title