Keypoint-based multi-label word segmentation and localization
US10878270B1 · kind B1 · utility
14Cited by
2References
20Claims
0Family size
Assignee
Inventors
Key dates
| Filing date | Jun 26, 2018 |
| Grant date | Dec 29, 2020 |
| Priority date | — |
| Expiry date | Oct 16, 2038 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06T2210/12
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
Techniques for keypoint-based multi-label word segmentation and localization are described. A machine learning model identifies bounding regions of text within an image, and then generates multiple channel matrices representing predicted keypoints of the text within the bounding regions. The keypoints can be used to rectify the corresponding graphical content from the image including the text to improve the ability to perform optical character recognition and identify the text. Line and word segmentation and localization can be performed together.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.