View source for 20240037969. RECOGNITION OF HANDWRITTEN TEXT VIA NEURAL NETWORKS simplified abstract (ABBYY Development Inc.)

=RECOGNITION OF HANDWRITTEN TEXT VIA NEURAL NETWORKS=


==Organization Name==

[[:Category:ABBYY Development Inc.|ABBYY Development Inc.]]


[[Category:ABBYY Development Inc.]]

==Inventor(s)==

[[:Category:Andrei Upshinskii of Kalinina (RU)|Andrei Upshinskii of Kalinina (RU)]][[Category:Andrei Upshinskii of Kalinina (RU)]]

==RECOGNITION OF HANDWRITTEN TEXT VIA NEURAL NETWORKS - A simplified explanation of the abstract==


This abstract first appeared for US patent application 20240037969 titled 'RECOGNITION OF HANDWRITTEN TEXT VIA NEURAL NETWORKS

==Simplified Explanation==

The abstract describes a system that receives an image of a line of text and segments it into multiple fragment images. The system then determines hypotheses to segment each fragment image into a plurality of grapheme images and calculates fragmentation confidence scores for each hypothesis. Based on the confidence scores, the system selects the hypothesis with the higher score and translates the grapheme images into symbols. Finally, the system assembles the symbols from each fragment image to reconstruct the original line of text.

* The system receives an image of a line of text and divides it into smaller fragment images.
* For each fragment image, the system generates hypotheses to segment it into grapheme images and calculates fragmentation confidence scores.
* The system compares the confidence scores and selects the hypothesis with the higher score.
* The selected hypothesis is used to translate the grapheme images into symbols.
* The symbols from each fragment image are combined to reconstruct the original line of text.

Potential applications of this technology:

* Optical character recognition (OCR) systems can benefit from improved text segmentation and reconstruction capabilities.
* Document scanning and digitization processes can be enhanced by accurately extracting text from images.
* Text-based search engines and indexing systems can improve their accuracy by correctly interpreting fragmented text.

Problems solved by this technology:

* Accurate segmentation of text in images can be challenging, especially when dealing with complex fonts or distorted text.
* Traditional OCR systems may struggle to correctly interpret fragmented text, leading to errors in text extraction and recognition.
* Reconstructing the original line of text from fragmented images can be time-consuming and error-prone without automated methods.

Benefits of this technology:

* Improved accuracy in text segmentation and reconstruction can enhance the overall performance of OCR systems.
* Efficient extraction of text from images can streamline document digitization processes and improve productivity.
* Enhanced interpretation of fragmented text can lead to more accurate search results and better indexing of textual content.


==Original Abstract Submitted==

in one embodiment, a system receives an image depicting a line of text. the system segments the image into two or more fragment images. for each of the two or more fragment images, the system determines a first hypothesis to segment the fragment image into a first plurality of grapheme images and a first fragmentation confidence score. the system determines a second hypothesis to segment the fragment image into a second plurality of grapheme images and a second fragmentation confidence score. the system determines that the first fragmentation confidence score is greater than the second fragmentation confidence score. the system translates the first plurality of grapheme images defined by the first hypothesis to symbols. the system assembles the symbols of each fragment image to derive the line of text.



[[Category:G06V30/148]]
[[Category:G06V30/413]]
[[Category:G06N3/04]]
[[Category:G06F18/214]]