Update app.py

app.py

@@ -17,9 +17,40 @@ pygame.mixer.init()
 yolo = YOLO("yolov8n.pt")
 
 # Streamlit app layout
-st.set_page_config(page_title="Assistive Vision App", layout="centered")
-st.title("Assistive Vision App for Visual Impairments")
-st.write("This application provides real-time object recognition and optional audio alerts.")
+st.set_page_config(page_title="Assistive Vision App", layout="wide")
+st.markdown(
+    """
+    <style>
+    body {
+        background-color: #f7f9fc;
+        font-family: "Arial", sans-serif;
+    }
+    .stButton>button {
+        background-color: #1a73e8;
+        color: white;
+        justify-content: center;
+        align-items: center;
+        border-radius: 10px;
+        padding: 10px;
+        margin: 5px;
+    }
+    .stCheckbox {
+        margin-top: 20px;
+    }
+    </style>
+    """,
+    unsafe_allow_html=True,
+)
+
+# Display welcome image
+welcome_image_path = "bismillah.png"  # Ensure this image exists in the script's directory
+if os.path.exists(welcome_image_path):
+    st.image(welcome_image_path, use_container_width=True, caption="Bismillah hir Rehman Ar Raheem")
+else:
+    st.warning("Welcome image not found! Please add 'bismillah.png' in the script directory.")
+
+st.title("Object Detection & Assistive Vision App for Visually Impaired People")
+st.write("This application provides real-time Object Recognition and Optional Audio Alerts.")
 
 # Directory to store temp audio files
 audio_temp_dir = "audio_temp_files"
@@ -30,8 +61,11 @@ if not os.path.exists(audio_temp_dir):
 stframe = st.empty()
 
 # User controls
-start_detection = st.button("Start Detection")
-stop_detection = st.button("Stop Detection")
+col1, col2 = st.columns(2)
+with col1:
+    start_detection = st.button("Start Detection")
+with col2:
+    stop_detection = st.button("Stop Detection")
 audio_activation = st.checkbox("Enable Audio Alerts", value=False)
 
 # Categories for audio alerts (hazardous objects or living things)
@@ -51,33 +85,24 @@ def play_audio_alert(label, position):
     ]
     caution_note = random.choice(phrases)
 
-    # Generate a unique audio filename using timestamp
     temp_file_path = os.path.join(audio_temp_dir, f"temp_{datetime.now().strftime('%Y%m%d_%H%M%S_%f')}.mp3")
 
-    # Generate audio
     tts = gTTS(caution_note)
     tts.save(temp_file_path)
 
     try:
-        # Load and play the audio using pygame
         pygame.mixer.music.load(temp_file_path)
         pygame.mixer.music.play()
 
-        # Create a new thread to handle file cleanup after playback
         def cleanup_audio_file():
-            while pygame.mixer.music.get_busy():  # Wait until audio finishes playing
+            while pygame.mixer.music.get_busy():
                 time.sleep(0.1)
-
-            # After music finishes, stop the music and try to remove the file
-            pygame.mixer.music.stop()  # Explicitly stop the music before cleanup
-
+            pygame.mixer.music.stop()
             try:
-                # Try to remove the file after playback is finished
                 os.remove(temp_file_path)
             except OSError as e:
                 print(f"Error deleting file {temp_file_path}: {e}")
 
-        # Start the cleanup thread
         threading.Thread(target=cleanup_audio_file, daemon=True).start()
 
     except Exception as e:
@@ -89,24 +114,20 @@ def process_frame(frame, audio_mode):
     results = yolo(frame)
     result = results[0]
 
-    detected_objects = {}  # Dictionary to store object labels and their positions
+    detected_objects = {}
     for box in result.boxes:
         x1, y1, x2, y2 = map(int, box.xyxy[0])
         label = result.names[int(box.cls[0])]
 
-        # Apply filtering logic for audio mode
         if audio_mode and label not in alert_categories:
             continue
 
-        # Determine position
         frame_center_x = frame.shape[1] // 2
         obj_center_x = (x1 + x2) // 2
         position = "left" if obj_center_x < frame_center_x else "right"
 
-        # Update detected objects with label and position
         detected_objects[label] = position
 
-        # Draw bounding boxes
         cv2.rectangle(frame, (x1, y1), (x2, y2), (0, 255, 0), 2)
         cv2.putText(
             frame,
@@ -135,26 +156,22 @@ if start_detection:
                     st.error("Failed to capture video. Please check your camera.")
                     break
 
-                # Process the frame based on whether audio alerts are enabled
                 detected_objects, processed_frame = process_frame(frame, audio_activation)
 
-                # Convert BGR to RGB for Streamlit
                 frame_rgb = cv2.cvtColor(processed_frame, cv2.COLOR_BGR2RGB)
                 stframe.image(frame_rgb, channels="RGB", use_container_width=True)
 
-                # Generate audio alerts for detected objects
                 if audio_activation:
                     current_time = datetime.now()
                     for label, position in detected_objects.items():
-                        # Check if enough time has passed since the last alert for this object
                         if (
                             label not in last_alert_time
                             or current_time - last_alert_time[label] > alert_cooldown
                         ):
                             play_audio_alert(label, position)
-                            last_alert_time[label] = current_time  # Update the last alert time
+                            last_alert_time[label] = current_time
 
-                time.sleep(0.1)  # Small delay for smoother updates
+                time.sleep(0.1)
 
     except Exception as e:
         st.error(f"An error occurred: {e}")