Api endpoints to call tesseract ocr functions
With docker:
docker run -p5005:5005 ieferrari/ocr_go_microservice
Get the text from an image url:
curl -X POST \
-d '{"msg": "https://pbs.twimg.com/media/EH-Pvo9WwAEKFwc?format=jpg&name=small"}' \
-H "Content-Type: application/json" \
http://127.0.0.1:5005/ocr_from_url
apt-get install automake ca-certificates g++ git libtool libleptonica-dev make pkg-config
git clone https://github.com/tesseract-ocr/tesseract.git
cd tesseract ./autogen.sh ./configure make sudo make install sudo ldconfig
https://notesalexp.org/tesseract-ocr/#tesseract_5.x sudo apt-get install tesseract-ocr
go get github.com/otiai10/gosseract/v2
wget https://github.com/tesseract-ocr/tessdata/raw/4.00/spa.traineddata tesseract --tessdata-dir . example.png outputbase -l spa --psm 3
On a 1 CPU, 1 GB RAM, vps server on Linode with Ubuntu 20.04.2 LTS
| service not running | running iddle | high load |
|---|---|---|
| 428 MB RAM | 530 MB RAM | 638 MB RAM |
| 4.4 % CPU | 5 % CPU | 60 % CPU |
Depending on your architecture, it may be more efficient to call tesseract from a wrapper in your preferred language. This container is an alternative if your team is having troubles installing the tesseract components for a specific language, or if you want a centralized ocr implementation in the first place.
Python example:
import pytesseract
from PIL import Image
pytesseract.pytesseract.tesseract_cmd ='/usr/local/bin/tesseract'
print(pytesseract.image_to_string(Image.open('./example.png'), lang='spa').replace("º", 'o'))