[ShikhaTan]

Scientific article PDFs have a common structure say Paper title, Abstract, Introduction, Proposed scheme, Experiments, Conclusion, References and so on. There will be sections and subsections for each.

How can I accept a PDF document as input and extract its physical structure. I want to get the physical structure of the PDF along with the data.

Eg: if the pdf document is like
Article Title
Abstract
sentences

Introduction
section-1
subsection-1
subsection-2
Proposed scheme
section-1
subsection-1

and so on….

I am coding in C#.

I want the logical structure of the document. Please remember that all scientific articles have a similar logical structure.

I have heard of some heuristic measures, but due to time constraint, I am looking for codes. Can anyone please help?

I am not allowed to use any publicly available libraries such as parsCit and so on. This is for my project work.

[Amit]

One thing to note is that all of these articles are actually written using Latex, with a specific template, e.g. have a look at this fir IEEE articles:

https://www.sharelatex.com/templates/journals/ieee-journal