Skip to content
TheChatArchiveThe Chat Archive
GPT 5.1· openaiHugging Face Frontier Reasoning Traces

Develop a Python function to perform optical character recognition (OCR) using Tesseract.

  1. User

    Develop a Python function to perform optical character recognition (OCR) using Tesseract.

  2. Assistant

    Developing OCR function

    The user is asking me to create a Python function for OCR using Tesseract without specifying a platform. I'm thinking of providing a snippet that uses the pytesseract library, and I’ll include installation instructions, explain how to configure the Tesseract path, and discuss handling different languages. It’s important to ensure the code is up-to-date for 2024 and handles errors effectively. I’ll also provide example usage in markdown format.Crafting OCR function

    I’ll provide installation commands for Tesseract on different operating systems: for Linux, it's apt install tesseract-ocr, and for macOS, brew install tesseract. For Windows, I’ll suggest installing the system binary and running pip install pytesseract pillow.

    Next, I’m crafting the function to perform OCR using pytesseract. It will accept either an image path or a PIL.Image and handle text extraction with optional language support. I’ll also include some example usage and a docstring for clarity! Below is a minimal, reusable Python function for OCR using Tesseract via the pytesseract wrapper.

    1. Install dependencies

    System Tesseract (required)

    • Ubuntu/Debian:
      sudo apt update
      sudo apt install tesseract-ocr
      
    • macOS (Homebrew):
      brew install tesseract
      
    • Windows:
      1. Download the Tesseract installer from:
        https://github.com/tesseract-ocr/tesseract/wiki/Downloads
      2. Install it, and note the installation path (e.g. C:\Program Files\Tesseract-OCR\tesseract.exe)

    Python packages

    pip install pytesseract pillow
    

    2. Python OCR function

    from typing import Union
    from PIL import Image
    import pytesseract
    
    
    # Uncomment and set this on Windows if pytesseract can't find Tesseract automatically:
    # pytesseract.pytesseract.tesseract_cmd = r"C:\Program Files\Tesseract-OCR\tesseract.exe"
    
    
    def perform_ocr(
        image: Union[str, Image.Image],
        lang: str = "eng",
        config: str | None = None
    ) -> str:
        """
        Run OCR on an image using Tesseract.
    
        Parameters
        ----------
        image : str or PIL.Image.Image
            Path to an image file (e.g., 'image.png') or a loaded PIL Image.
        lang : str, optional
            Tesseract language code(s), e.g. 'eng', 'deu', 'eng+spa'. Default is 'eng'.
        config : str, optional
            Extra Tesseract config, e.g. '--psm 6 --oem 3'.
    
        Returns
        -------
        str
            Extracted text from the image.
        """
        # Load image if a path is provided
        if isinstance(image, str):
            img = Image.open(image)
        elif isinstance(image, Image.Image):
            img = image
        else:
            raise TypeError("image must be a file path or PIL.Image.Image")
    
        # Optional simple pre-processing: convert to grayscale (often improves OCR)
        img = img.convert("L")
    
        # Run Tesseract OCR
        text = pytesseract.image_to_string(img, lang=lang, config=config or "")
        return text
    

    3. Example usage

    if __name__ == "__main__":
        # From image file path
        text = perform_ocr("sample_image.png", lang="eng")
        print(text)
    
        # With additional config (e.g., treat image as a single block of text)
        text2 = perform_ocr("scanned_page.jpg", lang="eng", config="--psm 6")
        print(text2)
    

    If you tell me your OS and typical image format (e.g., photos vs. scanned PDFs), I can adjust the function with better pre-processing steps for your case.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.