Upload 4 files

app.py

@@ -0,0 +1,130 @@
+import gradio as gr
+import cv2
+import numpy as np
+from ultralytics import YOLO
+import pytesseract
+import qrcode
+from pyzbar.pyzbar import decode
+import json
+from PIL import Image
+
+# Load YOLO models
+front_model = YOLO("front_cnic_model.pt")
+back_model = YOLO("back_cnic_model.pt")
+
+def preprocess_image(image):
+    # Convert PIL Image to numpy array
+    img = np.array(image)
+    # Convert RGB to BGR for OpenCV
+    img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
+    return img
+
+def extract_text(image, boxes):
+    # Perform OCR on detected regions
+    results = {}
+    for box in boxes:
+        x1, y1, x2, y2 = map(int, box.xyxy[0])
+        cls_name = front_model.names[int(box.cls)]
+        # Crop the region
+        roi = image[y1:y2, x1:x2]
+        # Convert to grayscale
+        gray = cv2.cvtColor(roi, cv2.COLOR_BGR2GRAY)
+        # Apply OCR
+        text = pytesseract.image_to_string(gray, config='--psm 6').strip()
+        if text:
+            results[cls_name] = text
+    return results
+
+def process_front_cnic(image):
+    try:
+        # Preprocess image
+        img = preprocess_image(image)
+        
+        # Run front CNIC detection
+        results = front_model.predict(img, conf=0.5)
+        
+        # Extract text from detected regions
+        extracted_info = extract_text(img, results[0].boxes)
+        
+        return extracted_info if extracted_info else {"error": "No text detected"}
+    except Exception as e:
+        return {"error": str(e)}
+
+def process_back_cnic(image):
+    try:
+        # Preprocess image
+        img = preprocess_image(image)
+        
+        # Run back CNIC detection
+        results = back_model.predict(img, conf=0.5)
+        
+        output = {}
+        # Process detected objects
+        for box in results[0].boxes:
+            cls_name = back_model.names[int(box.cls)]
+            x1, y1, x2, y2 = map(int, box.xyxy[0])
+            roi = img[y1:y2, x1:x2]
+            
+            if cls_name.lower() == "qr scan":
+                # Decode QR code
+                qr_result = decode(Image.fromarray(cv2.cvtColor(roi, cv2.COLOR_BGR2RGB)))
+                if qr_result:
+                    output["QR Scan"] = qr_result[0].data.decode('utf-8')
+                else:
+                    output["QR Scan"] = "No QR code detected"
+            
+            elif cls_name.lower() == "barcode":
+                # Decode barcode
+                barcode_result = decode(Image.fromarray(cv2.cvtColor(roi, cv2.COLOR_BGR2RGB)))
+                if barcode_result:
+                    output["Barcode"] = barcode_result[0].data.decode('utf-8')
+                else:
+                    output["Barcode"] = "No barcode detected"
+            
+            elif cls_name.lower() == "cnic":
+                # Extract CNIC number using OCR
+                gray = cv2.cvtColor(roi, cv2.COLOR_BGR2GRAY)
+                cnic_text = pytesseract.image_to_string(gray, config='--psm 6').strip()
+                output["CNIC"] = cnic_text if cnic_text else "No CNIC number detected"
+        
+        return output if output else {"error": "No objects detected"}
+    except Exception as e:
+        return {"error": str(e)}
+
+# Gradio Interface
+with gr.Blocks() as demo:
+    gr.Markdown("# CNIC Detection and Information Extraction")
+    
+    with gr.Tab("Front CNIC"):
+        front_input = gr.Image(type="pil", label="Upload Front CNIC Image")
+        front_output = gr.JSON(label="Extracted Information")
+        front_button = gr.Button("Process Front CNIC")
+    
+    with gr.Tab("Back CNIC"):
+        back_input = gr.Image(type="pil", label="Upload Back CNIC Image")
+        back_output = gr.JSON(label="Extracted Information")
+        back_button = gr.Button("Process Back CNIC")
+    
+    # Connect buttons to processing functions
+    front_button.click(
+        fn=process_front_cnic,
+        inputs=front_input,
+        outputs=front_output
+    )
+    
+    back_button.click(
+        fn=process_back_cnic,
+        inputs=back_input,
+        outputs=back_output
+    )
+
+# API endpoints
+api = gr.Interface(
+    fn=[process_front_cnic, process_back_cnic],
+    inputs=[gr.Image(type="pil"), gr.Image(type="pil")],
+    outputs=[gr.JSON(), gr.JSON()],
+    api_name="cnic_detection"
+)
+
+if __name__ == "__main__":
+    demo.launch()

back_cnic_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f697acb582e8549bf44611b44f0c0534d0b0a81350e3d4b6c151cc67d0a7c735
+size 6236899

front_cnic_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6823c4e300a2d43513912e69f7b8883fc94914bf81e8cf31926b38b0a588da53
+size 6258659

requirements.txt

@@ -0,0 +1,7 @@
+gradio==4.44.0 
+opencv-python==4.10.0.84 
+numpy==1.26.4 
+ultralytics==8.3.15 
+pytesseract==0.3.13 
+pyzbar==0.1.9 
+pillow==10.4.0