sundeep-news-scan/scratch_py/_ocr_page4.py

22 lines
943 B
Python

import paddleocr
from PIL import Image
from pathlib import Path
# Find the latest page 4 image
img_path = next(Path("output").glob("20260612_003429/pages/page_004.png"))
print(f"Running PaddleOCR on {img_path}")
ocr = paddleocr.PaddleOCR(lang='te')
result = ocr.predict(str(img_path))
print("OCR lines on Page 4 containing 'చెరువు' or 'మంత్రి' or numbers:")
if isinstance(result, list) and len(result) > 0:
det_result = result[0]
# Check if it has boxes and texts
texts = det_result.get('rec_texts', [])
boxes = det_result.get('dt_polys', [])
for i, txt in enumerate(texts):
if any(w in txt for w in ["చెరువు", "మంత్రి", "36", "ఏళ్ల", "ఉత్తమ్", "పల్లా", "సిద్ధపేట", "గూడెం"]):
box = boxes[i].tolist() if hasattr(boxes[i], 'tolist') else boxes[i]
print(f"Text: '{txt}' at box: {box}")