Back to results
US 2026/0120489 A1Published application
Context-aware ocr correction using llm
AssigneeSap SE
InventorTobias WELLER
Life of this patent
Estimating the term…
Abstract
Systems and methods include extraction of text data from an image, generation of a prompt including the extracted text data and indicating one or more portions of the extracted text data which represent handwritten text, input of the prompt to a text generation model, and reception of corrected text data from the text generation model in response to the prompt.
Claims20
A system comprising:
- a memory storing program code;
- and
- one or more processing units to execute the program code to cause the system to:
- acquire text data extracted from an image;
- generate a prompt including the extracted text data, instructions to correct the text data and indicating that the text data includes handwritten text;
- input the prompt to a text generation model;
- and
- receive corrected text data from the text generation model in response to the prompt.
5 more elements
A method comprising:
- extracting text data from an image;
- generating a prompt including the extracted text data and indicating one or more portions of the extracted text data which represent handwritten text;
- inputting the prompt to a text generation model;
- and
- receiving corrected text data from the text generation model in response to the prompt.
2 more elements
One or more non-transitory media storing program code executable by one or more processing units of a computing system to cause the computing system to:
- receive text data extracted from an image;
- generate a prompt including the extracted text data and indicating one or more portions of the extracted text data which represent handwritten text;
- input the prompt to a text generation model;
- and
- receive corrected text data from the text generation model in response to the prompt.
2 more elements
Dependencies are inferred from the claim text.