Spaces:
Sleeping
Sleeping
Create utils.py
Browse files
utils.py
ADDED
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import pytesseract
|
2 |
+
from PIL import Image
|
3 |
+
import re
|
4 |
+
import io
|
5 |
+
|
6 |
+
def extract_kyc_fields(image_bytes):
|
7 |
+
image = Image.open(io.BytesIO(image_bytes)).convert("RGB")
|
8 |
+
text = pytesseract.image_to_string(image)
|
9 |
+
|
10 |
+
# Extract Aadhaar number pattern XXXX-XXXX-XXXX or XXXX XXXX XXXX
|
11 |
+
aadhaar = re.search(r'\b\d{4}[\s-]?\d{4}[\s-]?\d{4}\b', text)
|
12 |
+
dob = re.search(r'\d{2}[\/\-]\d{2}[\/\-]\d{4}', text)
|
13 |
+
name_match = re.findall(r'(?i)(Name|Narne|NAMF)\s*[:\-]?\s*(.*)', text)
|
14 |
+
|
15 |
+
return {
|
16 |
+
"aadhaar_number": aadhaar.group(0) if aadhaar else "Not found",
|
17 |
+
"dob": dob.group(0) if dob else "Not found",
|
18 |
+
"name": name_match[0][1] if name_match else "Not found"
|
19 |
+
}
|