29 lines
406 B
Markdown
29 lines
406 B
Markdown
---
|
|
title: Convert a scanned pdf to text
|
|
tags:
|
|
- data
|
|
- pdf
|
|
- ocr
|
|
---
|
|
|
|
How to translate pdfs to text (results are very poor, and will need lots of corrections).
|
|
|
|
## Dependencies
|
|
|
|
Search for 'tesseract english' (or whatever language).
|
|
|
|
Arch: tesseract-data-eng and poppler-utils
|
|
|
|
## Script
|
|
|
|
```sh
|
|
pdftoppm -png *file*.pdf test
|
|
```
|
|
|
|
```sh
|
|
for x in *png; do
|
|
tesseract -l eng "$x" - >> out.txt
|
|
done
|
|
```
|
|
|