Skip to main content
Participant
September 19, 2025
Question

Adobe Extract API - Recognize Math Formula as single image

  • September 19, 2025
  • 1 reply
  • 207 views

Hello everyone,

i hope this is the right place to ask my specific question.

As questioned long time ago on Community Board: 

  1. https://community.adobe.com/t5/acrobat-services-api-discussions/math-equation-extraction-from-pdf/td-p/13033193 
  2. https://community.adobe.com/t5/acrobat-services-api-discussions/pdf-extract-api-math-image-extraction/td-p/14512088 

the question is still ongoing. How to extract math formulars as separate image from a PDF.

As example we can take Nr. 2 of the preceding question.

In the Example Project i don't found my anwer: https://github.com/adobe/PDFServices.NET.SDK.Samples 

Perhaps there other solutions, or it is not now implemented as I expected ...

Could anyone help me?

1 reply

Community Manager
September 19, 2025

Hi @gerald_0377,

Thanks for reaching out. I am moving your query to the correct forum where experts will be able to assist you further. In the meantime, you can check the following documents for more information:

https://adobe.ly/4msCOYD

https://adobe.ly/3VU7nLM

Hope this helps.

 

Regards,


^AN

Participant
October 8, 2025

Hi @Anshul_Nautiyal ,

unfortunally this dosent solve my problem, cause i had to know if the product adobe extract service could identify formulars and extract them from a pdf. Unfortunally its not sure if the user has the abillity to write the a math formular in the correct way e.g. Word, in some cases the are only pictures, e.g. copy & paste from a formular-writer ala google.

So thats the problem, there is the need to identify such images/regions as math formular and extract them as image to use them later in the pipline.

Regards

Gerald

Inspiring
October 8, 2025

@gerald_0377 

 

I'm not sure if this is what you are looking for....

 

We have built a product that remediates PDFs to make them accessible.  As part of that process we extract the images and generate an Alt-text description using an LLM.  As part of the process, we generate a report showing the image and the text in a side-by-side view.  Some of the documents we have processed have formulas, and here are a couple of the results.   Is this along the lines of what you are trying to achieve?