Update app.py

app.py

@@ -1,48 +1,49 @@
-
-
-from ultralytics import YOLO
 import streamlit as st
+import cv2
 import numpy as np
-from PIL import Image
-model=YOLO("best.pt")
+from ultralytics import YOLO
+
+# Load the YOLO model
+model = YOLO("best.pt")  # Ensure the path to your model is correct
+
 # Set the title of the app
-st.title("Fire Detection App")
+st.title("Live Fire Detection App")
 
-# Sidebar for input options
-input_option = st.sidebar.selectbox("Select Input Method", ["Upload Image"])
+# Create a placeholder for the video stream
+video_placeholder = st.empty()
 
-if input_option == "Upload Image":
-    # Upload Image
-    uploaded_file = st.file_uploader("Choose an Image", type=["jpg", "jpeg", "png"])
+# Function to perform live fire detection
+def detect_fire():
+    # Open a connection to the webcam (0 is the default camera)
+    cap = cv2.VideoCapture(0)
 
-    if uploaded_file is not None:
-        # Open and display the uploaded image
-        img = Image.open(uploaded_file)
-        st.image(img, caption='User Image', use_column_width=True)
-        
-        # Convert the image to a numpy array
-        img_np = np.array(img)
+    while True:
+        ret, frame = cap.read()
+        if not ret:
+            st.error("Failed to capture video.")
+            break
 
-        # Make predictions
-        results = model.predict(source=img_np, conf=0.5)  # Adjust confidence threshold as needed
+        # Convert the frame to RGB (as YOLO expects RGB input)
+        rgb_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
 
-        # Variable to check if fire is detected
-        fire_detected = False
+        # Make predictions on the current frame
+        results = model.predict(source=rgb_frame, conf=0.5)
 
-        # Draw bounding boxes on the image
+        # Draw bounding boxes on the frame
+        fire_detected = False
         for result in results:
             boxes = result.boxes.xyxy  # Bounding boxes
             for box in boxes:
                 x1, y1, x2, y2 = box[:4].astype(int)
-                img_np = cv2.rectangle(img_np, (x1, y1), (x2, y2), (0, 255, 0), 2)
+                rgb_frame = cv2.rectangle(rgb_frame, (x1, y1), (x2, y2), (0, 255, 0), 2)
 
                 # Check if the detected class is "fire" (adjust based on your model's class mapping)
                 class_id = int(box[5])  # Assuming class ID is at the 6th position
                 if class_id == 0:  # Replace 0 with the actual class ID for fire if different
                     fire_detected = True
 
-        # Show the resulting image with bounding boxes
-        st.image(img_np, caption='Detected Fire', use_column_width=True)
+        # Display the frame in the Streamlit app
+        video_placeholder.image(rgb_frame, channels="RGB", use_column_width=True)
 
         # Display message based on fire detection
         if fire_detected:
@@ -50,3 +51,14 @@ if input_option == "Upload Image":
         else:
             st.warning("No Fire Detected.")
 
+        # Break the loop if the user presses 'q'
+        if cv2.waitKey(1) & 0xFF == ord('q'):
+            break
+
+    # Release the webcam and close windows
+    cap.release()
+    cv2.destroyAllWindows()
+
+# Start the detection process
+if st.button("Start Live Detection"):
+    detect_fire()