ocrbychamod

Extract text from PNG and JPEG images using swappable OCR engines. Tesseract is the first built-in engine.

Prerequisites

The Tesseract OCR program must be installed on your machine. pip cannot install it for you.

• Windows: install the UB Mannheim build, then add C:\Program Files\Tesseract-OCR to your PATH
• macOS: brew install tesseract
• Ubuntu/Debian: sudo apt-get install tesseract-ocr

Check it works:

tesseract --version

Installation

pip install "ocrbychamod[tesseract]"

Usage

from ocrbychamod import extract_text

text = extract_text("receipt.png")
print(text)

If Tesseract is not on your PATH, or you want another language:

text = extract_text(
    "receipt.jpg",
    engine="tesseract",
    lang="eng",
    tesseract_cmd=r"C:\Program Files\Tesseract-OCR\tesseract.exe",
)

Supported formats: .png, .jpg, .jpeg.

Swappable engines

The core never depends on a specific OCR tool. Every engine implements the OCREngine interface, so you can plug in your own:

from ocrbychamod import OCREngine, extract_text, register_engine

class MyEngine(OCREngine):
    name = "my-engine"

    def extract_text(self, image):
        # image is a PIL.Image.Image
        return "recognized text"

# Option 1: pass an instance directly
extract_text("img.png", engine=MyEngine())

# Option 2: register it and select it by name
register_engine("my-engine", MyEngine)
extract_text("img.png", engine="my-engine")

List the registered engines:

from ocrbychamod import available_engines
print(available_engines())  # ['tesseract', ...]

Development

git clone https://github.com/Chamodpw2000/ocr_extract_pip.git
cd ocrbychamod
python -m venv .venv
.venv\Scripts\activate
pip install -e ".[tesseract]" pytest
pytest

License

MIT. See LICENSE.