app.py

import streamlit as st
import torch
import cv2
import numpy as np

# Title and description of the app
st.title("YOLOv5 Object Detection with Video Input")
st.write("Live object detection from your webcam using YOLOv5!")

# Load the pre-trained YOLOv5 model (COCO dataset)
@st.cache_resource
def load_model():
    return torch.hub.load('ultralytics/yolov5', 'yolov5s', pretrained=True)

model = load_model()

# Create a function to process video frames and apply YOLOv5
def process_frame(frame, model):
    # Perform inference
    results = model(frame)

    # Extract detections
    detections = results.pandas().xyxy[0]

    # Draw bounding boxes and labels on the frame
    for _, row in detections.iterrows():
        x1, y1, x2, y2 = int(row['xmin']), int(row['ymin']), int(row['xmax']), int(row['ymax'])
        label = f"{row['name']} {row['confidence']:.2f}"
        cv2.rectangle(frame, (x1, y1), (x2, y2), (255, 0, 0), 2)
        cv2.putText(frame, label, (x1, y1 - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (36, 255, 12), 2)

    return frame

# Start video capture
run_video = st.checkbox("Start Webcam")

if run_video:
    # Initialize the webcam
    cap = cv2.VideoCapture(0)  # 0 is the default camera

    if not cap.isOpened():
        st.error("Error: Could not open the webcam.")
    else:
        # Stream video
        stframe = st.empty()  # Placeholder for displaying video frames

        while run_video:
            ret, frame = cap.read()

            if not ret:
                st.error("Error: Failed to capture video.")
                break

            # Process the frame with YOLOv5
            processed_frame = process_frame(frame, model)

            # Convert BGR to RGB for Streamlit
            processed_frame = cv2.cvtColor(processed_frame, cv2.COLOR_BGR2RGB)

            # Display the frame in Streamlit
            stframe.image(processed_frame, channels="RGB", use_column_width=True)

        cap.release()
else:
    st.write("Enable the checkbox above to start the webcam.")