The process of dots.ocr for processing academic paper formulas is as follows:
- pass (a bill or inspection etc)Layout DetectionLocate the formula area in the document to generate accurate bounding box coordinates.
- utilizationContent Recognition ModuleConvert formulas to LaTeX format, maintaining the accuracy of mathematical notation and structure.
- In the JSON output result of the
formulafield to store the LaTeX code, while the Markdown file will start with the$...$Form inline formulas.
To improve the recognition rate, it is recommended to 1) ensure that the input image DPI is ≥200; 2) for dense formula regions can use theprompt_grounding_ocrwith manually labeled bounding boxes; 3) check the output for consecutive special characters (e.g.___) Whether post-processing correction is required.
This answer comes from the articledots.ocr: a unified visual-linguistic model for multilingual document layout parsingThe
May not be reproduced without permission:AI productivity tools " How to get the LaTeX representation of formulas when parsing academic papers?
English 
简体中文 
日本語 
Deutsch 
Português do Brasil