File size: 14,181 Bytes
1a7ee60 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 | """
Post-Processing and Key-Value Extraction
Combines NER predictions with regex patterns and heuristics
Implements confidence scoring and field validation
"""
import re
from typing import Dict, List, Optional, Tuple
import logging
from datetime import datetime
import numpy as np
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
class PostProcessor:
"""
Post-processing layer for IDP
Combines NER outputs with regex patterns and validation rules
"""
def __init__(self):
# Regex patterns for common fields
self.patterns = {
'invoice_number': [
r'INV[-\s]?\d+',
r'Invoice\s*#?\s*:?\s*([A-Z0-9\-]+)',
r'\b\d{6,}\b', # 6+ digits
],
'date': [
r'\d{1,2}[/-]\d{1,2}[/-]\d{2,4}',
r'\d{4}[/-]\d{1,2}[/-]\d{1,2}',
r'\d{1,2}\s+(?:Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)[a-z]*\s+\d{2,4}',
],
'amount': [
r'[₹$€£¥]\s*[\d,]+\.?\d*',
r'[\d,]+\.?\d*\s*[₹$€£¥]',
r'\b\d{1,3}(?:,\d{3})*(?:\.\d{2})?\b',
],
'gst_id': [
r'\d{2}[A-Z]{5}\d{4}[A-Z]{1}[A-Z\d]{1}[Z]{1}[A-Z\d]{1}', # Indian GST
r'[A-Z]{2}\d{9}', # VAT pattern
],
'email': [
r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b',
],
'phone': [
r'\+?\d{1,3}[-.\s]?\(?\d{1,4}\)?[-.\s]?\d{1,4}[-.\s]?\d{1,9}',
],
}
def process_document(
self,
document_type: str,
classification_confidence: float,
ocr_text: str,
ocr_boxes: List[Dict],
ner_entities: List[Dict]
) -> Dict:
"""
Main post-processing pipeline
Args:
document_type: Predicted document type (INVOICE, RECEIPT, FORM, OTHER)
classification_confidence: Confidence score from classifier
ocr_text: Full OCR extracted text
ocr_boxes: OCR bounding boxes with confidence
ner_entities: NER predicted entities
Returns:
Structured document with extracted fields
"""
logger.info(f"Post-processing {document_type} document")
# Extract fields based on document type
if document_type == 'INVOICE':
fields = self._extract_invoice_fields(ocr_text, ocr_boxes, ner_entities)
elif document_type == 'RECEIPT':
fields = self._extract_receipt_fields(ocr_text, ocr_boxes, ner_entities)
elif document_type == 'FORM':
fields = self._extract_form_fields(ocr_text, ocr_boxes, ner_entities)
elif document_type == 'BANK_STATEMENT':
fields = self._extract_bank_statement_fields(ocr_text, ocr_boxes, ner_entities)
else:
fields = self._extract_generic_fields(ocr_text, ocr_boxes, ner_entities)
# Build structured output
result = {
'document_type': document_type,
'classification_confidence': classification_confidence,
'fields': fields,
'raw_ocr_text': ocr_text[:1000], # First 1000 chars for debugging
}
return result
def _extract_invoice_fields(
self,
text: str,
boxes: List[Dict],
entities: List[Dict]
) -> Dict:
"""Extract invoice-specific fields"""
fields = {}
# Invoice number
invoice_num = self._extract_field(
'INVOICE_NUMBER',
text,
boxes,
entities,
self.patterns['invoice_number']
)
if invoice_num:
fields['invoice_number'] = invoice_num
# Date
date = self._extract_field(
'DATE',
text,
boxes,
entities,
self.patterns['date']
)
if date:
fields['date'] = self._normalize_date(date)
# Total amount
total = self._extract_field(
'TOTAL_AMOUNT',
text,
boxes,
entities,
self.patterns['amount']
)
if total:
fields['total_amount'] = self._normalize_amount(total)
# Tax amount
tax = self._extract_field(
'TAX_AMOUNT',
text,
boxes,
entities,
self.patterns['amount']
)
if tax:
fields['tax_amount'] = self._normalize_amount(tax)
# Vendor name
vendor = self._extract_field(
'VENDOR_NAME',
text,
boxes,
entities,
[]
)
if vendor:
fields['vendor_name'] = vendor
# Customer name
customer = self._extract_field(
'CUSTOMER_NAME',
text,
boxes,
entities,
[]
)
if customer:
fields['customer_name'] = customer
# GST/Tax ID
gst = self._extract_field(
'GST_ID',
text,
boxes,
entities,
self.patterns['gst_id']
)
if gst:
fields['gst_id'] = gst
return fields
def _extract_receipt_fields(
self,
text: str,
boxes: List[Dict],
entities: List[Dict]
) -> Dict:
"""Extract receipt-specific fields"""
# Similar to invoice but may have different priorities
return self._extract_invoice_fields(text, boxes, entities)
def _extract_form_fields(
self,
text: str,
boxes: List[Dict],
entities: List[Dict]
) -> Dict:
"""Extract form-specific fields"""
fields = {}
# Generic entity extraction
for entity in entities:
entity_type = entity['entity'].lower()
if entity_type not in fields:
fields[entity_type] = {
'value': entity['text'],
'confidence': entity['confidence'],
'bbox': [entity['start'], 0, entity['end'], 0] # Simplified
}
return fields
def _extract_bank_statement_fields(
self,
text: str,
boxes: List[Dict],
entities: List[Dict]
) -> Dict:
"""Extract bank statement specific fields"""
# Reuse invoice extraction logic for common fields like Date and Amount
return self._extract_invoice_fields(text, boxes, entities)
def _extract_generic_fields(
self,
text: str,
boxes: List[Dict],
entities: List[Dict]
) -> Dict:
"""Extract generic fields for unknown document types"""
return self._extract_form_fields(text, boxes, entities)
def _extract_field(
self,
entity_type: str,
text: str,
boxes: List[Dict],
entities: List[Dict],
fallback_patterns: List[str]
) -> Optional[Dict]:
"""
Extract a single field using NER + regex + heuristics
Returns:
Dictionary with value, confidence, and bbox, or None
"""
# Step 1: Check NER entities
for entity in entities:
if entity['entity'] == entity_type:
# Get OCR confidence for this text span
ocr_conf = self._get_ocr_confidence(entity, boxes)
# Combined confidence
combined_conf = (entity['confidence'] + ocr_conf) / 2
return {
'value': entity['text'],
'confidence': float(combined_conf),
'bbox': self._get_bbox_for_entity(entity, boxes),
'source': 'ner'
}
# Step 2: Fallback to regex patterns
if fallback_patterns:
for pattern in fallback_patterns:
matches = re.finditer(pattern, text, re.IGNORECASE)
for match in matches:
matched_text = match.group(0)
# Validate match
if self._validate_field(entity_type, matched_text):
# Find in OCR boxes
ocr_conf = self._find_in_boxes(matched_text, boxes)
return {
'value': matched_text,
'confidence': float(ocr_conf * 0.8), # Lower confidence for regex
'bbox': None,
'source': 'regex'
}
return None
def _validate_field(self, entity_type: str, value: str) -> bool:
"""Validate extracted field value"""
if not value or not value.strip():
return False
# Type-specific validation
if entity_type == 'DATE':
return len(value) >= 6 # Minimum date length
elif entity_type in ['TOTAL_AMOUNT', 'TAX_AMOUNT']:
return any(c.isdigit() for c in value)
elif entity_type == 'INVOICE_NUMBER':
return len(value) >= 3
return True
def _get_ocr_confidence(self, entity: Dict, boxes: List[Dict]) -> float:
"""Get average OCR confidence for entity text"""
entity_text = entity['text'].lower()
confidences = []
for box in boxes:
if box['text'].lower() in entity_text or entity_text in box['text'].lower():
confidences.append(box['confidence'])
return float(np.mean(confidences)) if confidences else 0.5
def _get_bbox_for_entity(self, entity: Dict, boxes: List[Dict]) -> Optional[List[float]]:
"""Get bounding box for entity from OCR boxes"""
entity_text = entity['text'].lower()
for box in boxes:
if entity_text in box['text'].lower():
return box['bbox']
return None
def _find_in_boxes(self, text: str, boxes: List[Dict]) -> float:
"""Find text in OCR boxes and return confidence"""
text_lower = text.lower()
for box in boxes:
if text_lower in box['text'].lower():
return float(box['confidence'])
return 0.5 # Default confidence
def _normalize_date(self, date_field: Dict) -> Dict:
"""Normalize date to standard format"""
date_str = date_field['value']
# Try to parse date
formats = [
'%d/%m/%Y', '%d-%m-%Y', '%Y-%m-%d', '%Y/%m/%d',
'%d/%m/%y', '%d-%m-%y',
'%d %B %Y', '%d %b %Y',
]
for fmt in formats:
try:
parsed = datetime.strptime(date_str, fmt)
date_field['value'] = parsed.strftime('%Y-%m-%d')
date_field['normalized'] = True
return date_field
except:
continue
# Could not parse, return as-is
date_field['normalized'] = False
return date_field
def _normalize_amount(self, amount_field: Dict) -> Dict:
"""Normalize amount to numeric value"""
amount_str = amount_field['value']
# Extract currency symbol
currency_symbols = {'₹': 'INR', '$': 'USD', '€': 'EUR', '£': 'GBP', '¥': 'JPY'}
currency = None
for symbol, code in currency_symbols.items():
if symbol in amount_str:
currency = code
amount_str = amount_str.replace(symbol, '')
break
# Remove commas and spaces
amount_str = amount_str.replace(',', '').replace(' ', '').strip()
# Extract numeric value
try:
numeric_value = float(amount_str)
amount_field['value'] = f"{numeric_value:.2f}"
amount_field['numeric_value'] = numeric_value
if currency:
amount_field['currency'] = currency
amount_field['normalized'] = True
except:
amount_field['normalized'] = False
return amount_field
if __name__ == "__main__":
# Example usage
processor = PostProcessor()
# Dummy data
ocr_text = """
INVOICE
Invoice #: INV-12345
Date: 28/11/2025
Total Amount: ₹ 12,500.00
Tax: ₹ 2,250.00
GST ID: 29ABCDE1234F1Z5
"""
ocr_boxes = [
{'text': 'INVOICE', 'bbox': [100, 50, 200, 80], 'confidence': 0.98},
{'text': 'INV-12345', 'bbox': [150, 100, 250, 120], 'confidence': 0.95},
{'text': '28/11/2025', 'bbox': [150, 130, 250, 150], 'confidence': 0.92},
{'text': '₹ 12,500.00', 'bbox': [200, 200, 300, 220], 'confidence': 0.96},
]
ner_entities = [
{'entity': 'INVOICE_NUMBER', 'text': 'INV-12345', 'confidence': 0.94, 'start': 30, 'end': 39},
{'entity': 'DATE', 'text': '28/11/2025', 'confidence': 0.89, 'start': 47, 'end': 57},
{'entity': 'TOTAL_AMOUNT', 'text': '₹ 12,500.00', 'confidence': 0.93, 'start': 75, 'end': 86},
]
result = processor.process_document(
document_type='INVOICE',
classification_confidence=0.96,
ocr_text=ocr_text,
ocr_boxes=ocr_boxes,
ner_entities=ner_entities
)
print("\nPost-processing result:")
print(json.dumps(result, indent=2))
|