'Form Recognizer - detection of superscripts and subscripts in tables
I plan to use Form Recognizer to detect and convert tables data from pdf documents. I already trained and used the custom model as not all of the tables were properly recognized automatically. I use Azure Function and FormRecognizerClient to get and convert data to required format. It works, but the issue is that these tables sometimes contains units in mathematical formulas with subscripts and superscripts.
The OCR doesn't recognize the power so f.ex 10^6 is recognized as 106, also some other signs are not recognized properly. Is there a way to improve this somehow? Maybe there is another Microsoft feature that I can use in parallel to detect things like this, or another tool to solve this issue?
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|