How is computer vision used for security applications?

Ultralytics YOLO models support a wide range of security applications through real-time object detection, tracking, and classification. Common use cases include PPE and workplace safety compliance, people and vehicle detection and tracking, perimeter and intrusion monitoring, access control and license-plate recognition, crowd and occupancy monitoring, threat and prohibited-object detection, and theft or anomaly detection. Because YOLO models can be trained on custom datasets, they’re typically fine-tuned for a specific environment, a factory floor, a parking structure, a retail store, rather than deployed as a generic, one-size-fits-all detector. Most security teams start with a single, well-defined detection task and expand into a broader platform combining tracking, behavior rules, and automated alerts as their needs grow.

Can YOLO be used for real-time security camera monitoring?

Yes. Ultralytics YOLO is built for real-time object detection, making it well suited for live CCTV and IP camera monitoring. It can detect people, vehicles, and objects across video streams with low latency, running on edge devices, on-premise servers, or existing NVR/DVR infrastructure, without needing to send footage to the cloud. This makes it a common choice for security teams that need instant alerts on intrusion, unauthorized access, or unsafe behavior rather than after-the-fact review.

What’s the difference between YOLO-based intrusion detection and traditional motion sensors?

Traditional motion sensors and light-curtain systems detect movement but can’t distinguish between a person, a vehicle, an animal, or debris, which leads to frequent false alarms. YOLO-based perimeter and intrusion detection uses computer vision to classify what triggered the event, apply zone-based rules, and analyze direction or loitering behavior. This reduces false positives significantly and lets security teams set more precise rules, such as alerting only on human intrusion into a restricted zone rather than any motion at all.

What license do I need to use YOLO for a commercial security product?

Ultralytics YOLO is available under the AGPL-3.0 open-source license or a commercial Enterprise License. If you’re building a commercial security or surveillance product, such as software you sell, license, or operate for customers, and you don’t want to open-source your application, you’ll need an Enterprise License rather than relying on AGPL. This is one of the most common questions from security vendors, since AGPL obligations can require sharing your full application source code if you distribute or host the software commercially.

Computer vision in security

From perimeter monitoring to theft detection, Ultralytics YOLO powers the next generation of security intelligence.

Trusted by the world's leading organizations

How Ultralytics YOLO tackles security

Real-time AI that works with your team

Built for security operations, Ultralytics YOLO streamlines workflows for precise, fast, and production-ready computer vision that runs on the cameras, VMS, and servers you already have.

Plug-and-play deployment: Works with your existing cameras and VMS, cutting integration time to days.
Detection accuracy: State-of-the-art real-time detection across objects, environments, and dynamic scenes.
Sub-5ms inference: Edge, cloud, or on-premise deployment with 20 export formats.
Production-ready in hours: Annotate, train, and deploy reducing time-to-market.

Request license

Proven results in security

From threat detection to access control and workplace safety, recent research shows what Ultralytics YOLO delivers for security teams.

Threat detection95.4%

Spot weapons in live video

Ultralytics YOLO26 detects weapons in surveillance footage at 95.4% mAP, and clears more than double the frames per second of earlier YOLO models on the same edge hardware.

Fierro Silva et al., Frontiers in Computer Science, 2026

Access control & LPR89%

Read plates and vehicles in one pass

An Ultralytics YOLO pipeline detects plates at 89% AP and turns each vehicle into a record of plate, color, make and model at 18 to 22 FPS.

Islas-Yañez et al., Sensors, 2026

Workplace safety7.3 ms

Check PPE compliance in real time

An Ultralytics YOLO11 safety monitor flags missing helmets, vests and glasses at 96.9% mAP in 7.3 ms a frame, down from 170 ms for earlier approaches on the same dataset.

Rahman et al., Computers, 2026

Try YOLO26 Inference

Drag and drop an image to see real-time object detection

Input Image

Upload an image or select an example to test yolo26n. Predict Docs

Live Demo

Drop an image here

or click to browse

Supports JPEG, PNG, WebP, AVIF, HEIC, JP2, TIFF, BMP & more (max 10MB)

Or try an example

Confidence

0.25

IoU

0.70

Image Size

640

API Docs

Deploy your model to get a dedicated inference endpoint with API key authentication — then replace the placeholder URL and key below with your deployment values. Deploy docs

Note: This is example code. Deploy your model first to get your endpoint URL and API key.

import requests

# Replace with your deployment URL and API key
url = "https://your-deployment-url.run.app/predict"
api_key = "YOUR_API_KEY"

# Optional inference parameters (conf, iou, imgsz)
args = {"conf": 0.25, "iou": 0.7, "imgsz": 640}

with open("image.jpg", "rb") as f:
    response = requests.post(
        url,
        headers={"Authorization": f"Bearer {api_key}"},
        data=args,
        files={"file": f},
    )

print(response.json())

Vision AI for every stage in security

Purpose-built solutions for every stage of your production process.

Perimeter monitoring

Intrusion monitoring with Ultralytics YOLO

Leverage Ultralytics YOLO to know the moment someone enters a restricted area. Define zones directly on your existing feeds and get alerted on intrusion, tailgating, or unauthorized access.

Zone-based detection: Monitor restricted areas and get alerted when crossed.
Access misuse: Catch repeated use of a single credential or badge.
Loitering & direction analysis: Flag suspicious dwell time or movement against expected flow.