File size: 4,317 Bytes
40c3856 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 |
import os
import uuid
from flask import Flask, request, jsonify
from flask_cors import CORS
from werkzeug.utils import secure_filename
import torch
from transformers import DonutProcessor, VisionEncoderDecoderModel
from PIL import Image
import cv2
import numpy as np
app = Flask(__name__)
CORS(app)
# Configure upload folder
UPLOAD_FOLDER = 'uploads'
ALLOWED_EXTENSIONS = {'png', 'jpg', 'jpeg', 'pdf', 'tif', 'tiff'}
app.config['UPLOAD_FOLDER'] = UPLOAD_FOLDER
app.config['MAX_CONTENT_LENGTH'] = 16 * 1024 * 1024 # 16MB max upload
# Create uploads directory if it doesn't exist
os.makedirs(UPLOAD_FOLDER, exist_ok=True)
# Load OCR model - Microsoft's Donut model
processor = DonutProcessor.from_pretrained("microsoft/donut-base")
model = VisionEncoderDecoderModel.from_pretrained("microsoft/donut-base")
# Move model to GPU if available
device = "cuda" if torch.cuda.is_available() else "cpu"
model.to(device)
def allowed_file(filename):
return '.' in filename and \
filename.rsplit('.', 1)[1].lower() in ALLOWED_EXTENSIONS
def preprocess_image(image_path):
# Open image with PIL
image = Image.open(image_path).convert("RGB")
# Basic enhancement for better OCR results
# Convert to OpenCV format for preprocessing
img = np.array(image)
img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
# Apply adaptive thresholding to handle varying lighting conditions
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
thresh = cv2.adaptiveThreshold(gray, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,
cv2.THRESH_BINARY, 11, 2)
# Convert back to PIL
enhanced_image = Image.fromarray(cv2.cvtColor(thresh, cv2.COLOR_GRAY2RGB))
return enhanced_image
def perform_ocr(image_path):
# Preprocess the image
image = preprocess_image(image_path)
# Prepare image for the model
pixel_values = processor(image, return_tensors="pt").pixel_values.to(device)
# Generate text
task_prompt = "<s_ocr>"
decoder_input_ids = processor.tokenizer(task_prompt, add_special_tokens=False, return_tensors="pt").input_ids.to(device)
outputs = model.generate(
pixel_values,
decoder_input_ids=decoder_input_ids,
max_length=model.decoder.config.max_position_embeddings,
early_stopping=True,
pad_token_id=processor.tokenizer.pad_token_id,
eos_token_id=processor.tokenizer.eos_token_id,
use_cache=True,
num_beams=5,
bad_words_ids=[[processor.tokenizer.unk_token_id]],
return_dict_in_generate=True,
)
# Decode generated text
sequence = processor.batch_decode(outputs.sequences)[0]
sequence = sequence.replace(processor.tokenizer.eos_token, "").replace(processor.tokenizer.pad_token, "")
sequence = sequence.replace("<s>", "").replace("</s>", "").replace("<s_ocr>", "").replace("</s_ocr>", "")
return sequence.strip()
@app.route('/ocr', methods=['POST'])
def ocr():
# Check if a file was uploaded
if 'file' not in request.files:
return jsonify({'error': 'No file part'}), 400
file = request.files['file']
# Check if filename is empty
if file.filename == '':
return jsonify({'error': 'No selected file'}), 400
# Check if file type is allowed
if file and allowed_file(file.filename):
# Create a unique filename
filename = str(uuid.uuid4()) + '_' + secure_filename(file.filename)
file_path = os.path.join(app.config['UPLOAD_FOLDER'], filename)
# Save the file
file.save(file_path)
try:
# Perform OCR
extracted_text = perform_ocr(file_path)
# Clean up the file if needed
# os.remove(file_path)
return jsonify({
'success': True,
'text': extracted_text
})
except Exception as e:
return jsonify({
'success': False,
'error': str(e)
}), 500
else:
return jsonify({'error': 'File type not allowed'}), 400
@app.route('/health', methods=['GET'])
def health_check():
return jsonify({'status': 'healthy'}), 200
if __name__ == '__main__':
app.run(host='0.0.0.0', port=5000, debug=False |