Update app.py

app.py

@@ -2,15 +2,13 @@ import cv2
 import streamlit as st
 import numpy as np
 from PIL import Image
+import time
 
 # Load the pre-trained Haar Cascade face detector
 face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')
 
+# Define a function to detect faces in a frame
 def detect_faces(frame):
-    """
-    Detect faces in the frame.
-    Returns the frame with bounding boxes drawn around detected faces.
-    """
     # Convert the frame to grayscale (Haar Cascade works on grayscale images)
     gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
 
@@ -26,23 +24,34 @@ def detect_faces(frame):
 # Streamlit UI for the app
 st.title("Real-Time Face Detection")
 
-# Capture the video from the webcam
-camera = st.camera_input("Capture a photo")
+# Create a container for the webcam video feed
+video_placeholder = st.empty()
 
-# Process the webcam image if available
-if camera:
-    # Convert the camera image into a numpy array
-    img = Image.open(camera)
-    img_array = np.array(img)
-    
-    # Convert the image to a format OpenCV can process (BGR)
-    img_bgr = cv2.cvtColor(img_array, cv2.COLOR_RGB2BGR)
+# Start the webcam feed
+cap = cv2.VideoCapture(0)  # 0 is the default webcam device
 
-    # Detect faces in the image
-    result_frame = detect_faces(img_bgr)
+if not cap.isOpened():
+    st.error("Error: Could not access the webcam.")
+else:
+    # Start capturing video frames
+    while True:
+        ret, frame = cap.read()
+
+        if not ret:
+            st.error("Failed to grab frame.")
+            break
+
+        # Detect faces in the current frame
+        result_frame = detect_faces(frame)
+
+        # Convert BGR (OpenCV format) to RGB (for Streamlit display)
+        result_frame_rgb = cv2.cvtColor(result_frame, cv2.COLOR_BGR2RGB)
+
+        # Display the frame in Streamlit (dynamically updating the image)
+        video_placeholder.image(result_frame_rgb, channels="RGB", use_column_width=True)
 
-    # Convert result frame back to RGB (for displaying in Streamlit)
-    result_frame_rgb = cv2.cvtColor(result_frame, cv2.COLOR_BGR2RGB)
+        # Add a small delay to control the frame rate (optional)
+        time.sleep(0.03)  # This gives roughly 30 FPS
 
-    # Display the result in Streamlit
-    st.image(result_frame_rgb, caption="Detected Faces", use_column_width=True)
+# Release the webcam when the stream ends
+cap.release()