Run Real-Time Object Detection with YOLOv8 and Python

1. 1
   Install Dependencies

   Set up your Python environment by installing the core `ultralytics` library, which includes YOLOv8, and `opencv-python` for handling video streams from your webcam or video files.
2. 2
   Load a Pre-trained Model

   Instantiate a YOLO model. The `ultralytics` library automatically downloads the specified pre-trained model weights the first time you use them. We'll use 'yolov8n.pt', a small and fast model ideal for real-time applications.
3. 3
   Run Detection on a Single Image

   Perform a test inference on a single image to verify your setup. The `predict` method runs the detection and returns a list of results. We'll save the annotated image to a file to see the output.
4. 4
   Process Detection Results

   The power of the Object Detection Setup is accessing the structured data. Iterate through the `boxes` attribute of a result to get coordinates (xyxy), confidence scores, and class IDs for each detected object.
5. 5
   Launch Real-Time Webcam Feed

   Use OpenCV to capture frames from your webcam, pass each frame to the YOLO model for inference, and display the annotated video stream in real-time. The starter code below provides a complete implementation of this.

import cv2
from ultralytics import YOLO
import json

# Load the YOLOv8 model
model = YOLO('yolov8n.pt')

# Open the webcam
cap = cv2.VideoCapture(0)

# Check if the webcam is opened correctly
if not cap.isOpened():
    raise IOError("Cannot open webcam")

print("Webcam opened. Press 'q' to quit.")

while True:
    # Read a frame from the webcam
    ret, frame = cap.read()
    if not ret:
        break

    # Run YOLOv8 inference on the frame
    results = model(frame)

    # --- Process Detections for JSON Output ---
    detections = []
    for r in results:
        for box in r.boxes:
            detection = {
                'class_id': int(box.cls.item()),
                'class_name': model.names[int(box.cls.item())],
                'confidence': float(box.conf.item()),
                'box': [float(coord) for coord in box.xyxy[0].tolist()]
            }
            detections.append(detection)
    
    # Print detections as a JSON string for the current frame
    if detections:
        print(json.dumps(detections, indent=2))

    # Visualize the results on the frame
    annotated_frame = results[0].plot()

    # Display the annotated frame
    cv2.imshow('YOLOv8 Real-Time Detection', annotated_frame)

    # Break the loop if 'q' is pressed
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

# Release the webcam and close the display window
cap.release()
cv2.destroyAllWindows()
print("Webcam feed closed.")