Update app.py

app.py

@@ -1,30 +1,17 @@
 import gradio as gr
-import pytesseract
-from PIL import Image
-import json
+from utils import extract_kyc_fields
 
-def extract_fields(image):
-    try:
-        raw_text = pytesseract.image_to_string(image)
-        lines = raw_text.split('\n')
-        result = {}
+def process_document(file):
+    result = extract_kyc_fields(file.read())
+    return result
 
-        for line in lines:
-            line = line.strip()
-            if ':' in line:
-                key, value = line.split(':', 1)
-                result[key.strip()] = value.strip()
-            elif len(line.split()) >= 2:
-                result["line_" + str(len(result))] = line
+iface = gr.Interface(
+    fn=process_document,
+    inputs=gr.File(label="Upload Aadhaar/PAN image", type="binary"),
+    outputs=gr.JSON(label="Extracted KYC Fields"),
+    title="Smart KYC OCR",
+    description="Upload an Aadhaar or PAN card image to extract structured KYC fields using Tesseract OCR"
+)
 
-        return json.dumps(result, indent=2)
-    except Exception as e:
-        return {"error": str(e)}
-
-gr.Interface(
-    fn=extract_fields,
-    inputs=gr.Image(type="pil"),
-    outputs="json",
-    title="Smart KYC OCR (Tesseract)",
-    description="Upload Aadhaar or PAN image to extract key-value KYC fields using Tesseract OCR."
-).launch()
+if __name__ == "__main__":
+    iface.launch()