YOLOv8 can run object detection on images with a pre-trained model. YOLO (You Only Look Once) is a well-known object-detection algorithm, and YOLOv8 is designed for fast inference and accurate results. The ultralytics library provides the Python interface used below.

Requirements

Before you begin, make sure you have:

  • Python
  • The ultralytics library

If it is not installed yet, install it with pip:

pip install ultralytics

Run YOLOv8 on an image

This example loads a pre-trained YOLOv8 model, runs it on one image, and exposes the detection results:

from ultralytics import YOLO

# Load the default pre-trained YOLOv8 model
model = YOLO("yolov8n.pt")

# Run detection on an image
results = model(["image/car.jpg"])

# Iterate over the detection results
for result in results:
    boxes = result.boxes      # bounding boxes
    masks = result.masks      # segmentation masks, when supported by the model
    keypoints = result.keypoints  # pose keypoints
    probs = result.probs      # classification probabilities
    obb = result.obb          # oriented bounding boxes

    # Display the result
    result.show()

    # Save the result to a file
    result.save(filename="result.jpg")

What the example returns

  • Load the model → YOLO("yolov8n.pt") loads the pre-trained YOLOv8 Nano model. If it is not available locally, Ultralytics downloads it automatically.
  • Detect objects → model(["image/car.jpg"]) runs inference on the image. Replace the path with any image you want to analyze.
  • Read the results → the result can contain bounding boxes, segmentation masks, pose keypoints, classification probabilities, and other task-specific output.
  • Output → display the annotated result with result.show() or save it with result.save().

Next steps

The same model can process a batch of image paths or serve as the starting point for real-time detection from a camera. Replace image/car.jpg with an image available in your environment and inspect the returned result fields for the task you need.

AI