Files
railseek6/test_paddleocr_cpu.py

58 lines
2.1 KiB
Python

#!/usr/bin/env python3
"""
Test script for PaddleOCR in CPU mode
"""
import os
import sys
print('🔍 Testing PaddleOCR with corrected parameters...')
try:
# Test PaddleOCR with correct parameter names
from paddleocr import PaddleOCR
print('✅ PaddleOCR imported successfully')
# Initialize OCR with correct parameters (no use_gpu parameter in this version)
ocr = PaddleOCR(use_textline_orientation=True, lang='en')
print('✅ PaddleOCR initialized successfully')
print('💡 Note: GPU mode requires cuDNN installation for CUDA 12.9')
# Test with a simple image path
test_image = 'test_ocr_image.png'
if os.path.exists(test_image):
print(f'✅ Test image found: {test_image}')
result = ocr.ocr(test_image, cls=True)
if result and len(result) > 0:
print('✅ OCR processing successful')
print(f' Detected {len(result[0])} text elements')
# Show first few results
for i, line in enumerate(result[0][:3]):
text = line[1][0]
confidence = line[1][1]
print(f' {i+1}. \"{text}\" (confidence: {confidence:.2f})')
else:
print('❌ OCR processing returned no results')
else:
print(f'⚠️ Test image not found: {test_image}')
print('💡 Testing OCR initialization only')
except Exception as e:
print(f'❌ PaddleOCR test failed: {e}')
import traceback
traceback.print_exc()
print('\n🔍 Checking if OCR PDF exists...')
ocr_pdf_path = 'inputs/__enqueued__/ocr.pdf'
if os.path.exists(ocr_pdf_path):
print(f'✅ OCR PDF found: {ocr_pdf_path}')
print('💡 Ready to test OCR PDF upload')
else:
print(f'❌ OCR PDF not found: {ocr_pdf_path}')
print('💡 Please ensure ocr.pdf is in inputs/__enqueued__/ directory')
print('\n🔍 Summary:')
print('✅ PaddleOCR works in CPU mode')
print('❌ GPU acceleration requires cuDNN for CUDA 12.9')
print('💡 Download cuDNN from: https://developer.nvidia.com/cudnn')
print('💡 Extract to: C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.9\\bin\\')