Create app.py

app.py

@@ -0,0 +1,47 @@
+from paddleocr import PaddleOCR
+import gradio as gr
+import re
+
+# Initialize PaddleOCR
+ocr = PaddleOCR(use_angle_cls=True, lang='en')
+
+def extract_kyc_fields(image_path):
+    result = ocr.ocr(image_path, cls=True)
+    text_lines = [line[1][0] for line in result[0]]
+
+    output = {
+        "aadhaar_number": None,
+        "name": None,
+        "dob": None
+    }
+
+    # Aadhaar number
+    for text in text_lines:
+        match = re.search(r"\d{4}[\s-]?\d{4}[\s-]?\d{4}", text)
+        if match:
+            output["aadhaar_number"] = match.group().replace(" ", "-")
+            break
+
+    # DOB
+    for text in text_lines:
+        match = re.search(r"(\d{2,4})[/-](\d{1,2})[/-](\d{2,4})", text)
+        if match:
+            output["dob"] = match.group()
+            break
+
+    # Name
+    for text in text_lines:
+        if text.isupper() and not any(char.isdigit() for char in text):
+            output["name"] = text.title()
+            break
+
+    return output  # Return dictionary directly (key-value)
+
+# Gradio interface
+gr.Interface(
+    fn=extract_kyc_fields,
+    inputs=gr.Image(type="filepath", label="Upload Aadhaar/PAN Card"),
+    outputs=gr.JSON(label="Extracted KYC Fields"),  # Key-value format
+    title="🧠 Smart KYC OCR",
+    description="Upload Aadhaar/PAN card image and get extracted Name, Aadhaar number, and DOB as key-value output."
+).launch()