Python OCR PDF Extraction

Tesseract Installation

• ocr2.py
  

1. #!/usr/bin/python3
2. #
3. # Python OCR PDF Extraction
4. # https://github.com/tesseract-ocr/tesseract
5. #
6. # sudo apt install tesseract-ocr
7. # sudo apt install libtesseract-dev
8. # pip install pytesseract PyPDF2 pdfplumber opencv-python pillow
9. # pip install pdf2image
10. # sudo apt-get install poppler-utils
11. # sudo apt-get install tesseract-ocr-chi-sim  # Simplified Chinese
12. # sudo apt-get install tesseract-ocr-chi-tra  # Traditional Chinese
13. # tesseract --list-langs
15. import pytesseract
16. from pdf2image import convert_from_path
17. from PyPDF2 import PdfReader
18. import cv2
19. import numpy as np
20. from PIL import Image
22. # Path to Tesseract executable (update to match your system)
23. pytesseract.pytesseract.tesseract_cmd = '/usr/bin/tesseract'
25. def preprocess_image(pil_image):
26.     """
27.     Preprocesses an image for OCR using OpenCV.
28.     Converts to grayscale, applies thresholding.
29.     """
30.     # Convert PIL image to OpenCV format
31.     open_cv_image = np.array(pil_image)
32.     # Convert RGB to BGR (OpenCV default format)
33.     open_cv_image = cv2.cvtColor(open_cv_image, cv2.COLOR_RGB2BGR)
34.     # Convert to grayscale
35.     gray_image = cv2.cvtColor(open_cv_image, cv2.COLOR_BGR2GRAY)
36.     # Apply binary thresholding
37.     _, thresh_image = cv2.threshold(gray_image, 128, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
38.     return thresh_image
40. def extract_text_from_pdf(pdf_path):
41.     # First try extracting text from the PDF directly
42.     reader = PdfReader(pdf_path)
43.     text = ""
44.     for page in reader.pages:
45.         text += page.extract_text() or ""
47.     # If no text is extracted, assume it's a scanned PDF and use OCR
48.     if not text.strip():
49.         images = convert_from_path(pdf_path)
50.         for image in images:
51.             # Preprocess image for better OCR results
52.             preprocessed_image = preprocess_image(image)
53.             # Convert OpenCV image back to PIL format for Tesseract
54.             pil_image = Image.fromarray(preprocessed_image)
55.             # Perform OCR
56.             text += pytesseract.image_to_string(pil_image, lang='chi_sim')
58.     return text
60. # Example usage
61. pdf_path = "scan_2025-01-02_09.31.pdf"
62. extracted_text = extract_text_from_pdf(pdf_path)
63. print(extracted_text)

复制代码

免责声明：如果侵犯了您的权益，请联系站长，我们会及时删除侵权内容，谢谢合作！更多信息从访问主页：qidao123.com:ToB企服之家，中国第一个企服评测及商务社交产业平台。