aiopytesseract

A Python asyncio wrapper for Tesseract-OCR.

Installation

Install and update using pip:

pip install aiopytesseract

Usage

List all available languages by Tesseract installation

import aiopytesseract

await aiopytesseract.languages()
await aiopytesseract.get_languages()

Tesseract version

import aiopytesseract

await aiopytesseract.tesseract_version()
await aiopytesseract.get_tesseract_version()

Tesseract parameters

import aiopytesseract

await aiopytesseract.tesseract_parameters()

Confidence only info

import aiopytesseract

await aiopytesseract.confidence("tests/samples/file-sample_150kB.png")

Deskew info

import aiopytesseract

await aiopytesseract.deskew("tests/samples/file-sample_150kB.png")

Extract text from an image: locally or bytes

from pathlib import Path

import aiopytesseract

await aiopytesseract.image_to_string("tests/samples/file-sample_150kB.png")
await aiopytesseract.image_to_string(
	Path("tests/samples/file-sample_150kB.png").read_bytes(), dpi=220, lang='eng+por'
)

Box estimates

from pathlib import Path

import aiopytesseract

await aiopytesseract.image_to_boxes("tests/samples/file-sample_150kB.png")
await aiopytesseract.image_to_boxes(Path("tests/samples/file-sample_150kB.png")

Boxes, confidence and page numbers

from pathlib import Path

import aiopytesseract

await aiopytesseract.image_to_data("tests/samples/file-sample_150kB.png")
await aiopytesseract.image_to_data(Path("tests/samples/file-sample_150kB.png")

Information about orientation and script detection

from pathlib import Path

import aiopytesseract

await aiopytesseract.image_to_osd("tests/samples/file-sample_150kB.png")
await aiopytesseract.image_to_osd(Path("tests/samples/file-sample_150kB.png")

Generate a searchable PDF

from pathlib import Path

import aiopytesseract

await aiopytesseract.image_to_pdf("tests/samples/file-sample_150kB.png")
await aiopytesseract.image_to_pdf(Path("tests/samples/file-sample_150kB.png")

Generate HOCR output

from pathlib import Path

import aiopytesseract

await aiopytesseract.image_to_hocr("tests/samples/file-sample_150kB.png")
await aiopytesseract.image_to_hocr(Path("tests/samples/file-sample_150kB.png")

Multi ouput

from pathlib import Path

import aiopytesseract

async with aiopytesseract.run(
	Path('tests/samples/file-sample_150kB.png').read_bytes(),
	'output',
	'alto tsv txt'
) as resp:
	# will generate (output.xml, output.tsv and output.txt)
	print(resp)
	alto_file, tsv_file, txt_file = resp

Config variables

from pathlib import Path

import aiopytesseract

async with aiopytesseract.run(
	Path('tests/samples/text-with-chars-and-numbers.png').read_bytes(),
	'output',
	'alto tsv txt'
	config=[("tessedit_char_whitelist", "0123456789")]
) as resp:
	# will generate (output.xml, output.tsv and output.txt)
	print(resp)
	alto_file, tsv_file, txt_file = resp

from pathlib import Path

import aiopytesseract

await aiopytesseract.image_to_string(
	"tests/samples/text-with-chars-and-numbers.png",
	config=[("tessedit_char_whitelist", "0123456789")]
)

await aiopytesseract.image_to_string(
	Path("tests/samples/text-with-chars-and-numbers.png").read_bytes(),
	dpi=220,
	lang='eng+por',
	config=[("tessedit_char_whitelist", "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ")]
)

For more details on Tesseract best practices and the aiopytesseract, see the folder: docs.

Examples

If you want to test aiopytesseract easily, can you use some options like:

• docker/docker-compose
• streamlit

Docker / docker-compose

After clone this repo run the command below:

docker-compose up -d

streamlit app

For this option it's necessary first install aiopytesseract and streamlit, after execute:

# remote option:
streamlit run https://github.com/amenezes/aiopytesseract/blob/master/examples/streamlit/app.py

# local option:
streamlit run examples/streamlit/app.py

note: The streamlit example need python >= 3.10

Links

• License: Apache License
• Code: https://github.com/amenezes/aiopytesseract
• Issue tracker: https://github.com/amenezes/aiopytesseract/issues
• Docs: https://github.com/amenezes/aiopytesseract