Preset: NVIDIA Jetson {#jetson}

Real-time gun detection using Frigate NVR with TensorRT GPU acceleration on NVIDIA Jetson.

What you'll get:

• Real-time gun detection with bounding boxes and confidence scores
• Web dashboard for live monitoring and event review
• Automatic recording and snapshots when guns are detected
• MQTT integration for alerts and automation

Requirements: Docker + NVIDIA Container Toolkit on target device

Step 1: Initialize Cameras {#init_cameras_jetson type=manual required=false}

Set up your IP cameras and obtain RTSP stream URLs.

Wiring

1. Connect your IP camera to the same network as the deployment device
2. Find the camera's IP address (check your router's DHCP client list, or use the manufacturer's tool)
3. Access the camera's web management interface (usually http://<camera-ip>)
4. Enable RTSP streaming in the camera settings (usually enabled by default)
5. Note down the RTSP URL for your camera

Common RTSP URL formats:

• Hikvision: rtsp://admin:password@<ip>:554/Streaming/Channels/101 (main stream) or /102 (sub stream)
• Dahua: rtsp://admin:password@<ip>:554/cam/realmonitor?channel=1&subtype=0 (main) or &subtype=1 (sub)
• Generic ONVIF: Use the camera's ONVIF discovery tool to find the stream URL

Tip: Use the sub stream (lower resolution) for detection to reduce CPU/GPU load, and the main stream for recording.

You can test your RTSP URL with VLC: Media > Open Network Stream > paste the URL.

Troubleshooting

Step 2: Deploy Frigate {#deploy_frigate_jetson type=docker_deploy required=true config=devices/jetson.yaml}

Deploy Frigate NVR with TensorRT-accelerated gun detection to your NVIDIA Jetson.

Target: Remote Deployment {#jetson_remote type=remote config=devices/jetson.yaml default=true}

Deploy to a remote NVIDIA Jetson device via SSH.

Wiring

1. Connect Jetson to the same network as your computer
2. Enter the Jetson's IP address and SSH credentials
3. Optionally enter your RTSP camera URLs (up to 2 cameras)
4. Click Deploy to start installation

Note: First startup takes 5-10 minutes — the system downloads and compiles the gun detection model for TensorRT optimization.

Deployment Complete

1. Open http://<device-ip>:5000 in your browser
2. You'll see two demo cameras detecting guns in sample videos
3. If you provided RTSP URLs, your cameras will also appear with gun detection enabled

To modify camera configuration later, SSH into the device and edit:

cd ~/gun-detection-frigate
nano config/config.yml
docker compose restart

Troubleshooting

Target: Local Deployment {#jetson_local type=local config=devices/jetson_local.yaml}

Deploy directly on the current machine (requires NVIDIA GPU).

Wiring

1. Ensure Docker and NVIDIA Container Toolkit are installed
2. Click Deploy to start installation

Note: First startup takes 5-10 minutes for TensorRT model compilation.

Deployment Complete

1. Open http://localhost:5000 in your browser
2. You'll see two demo cameras detecting guns in sample videos

To add RTSP cameras, edit the config file and restart:

cd ~/gun-detection-frigate
nano config/config.yml
docker compose restart

Troubleshooting

Preset: reComputer R2000 + Hailo {#r2000_hailo}

Real-time gun detection using Frigate NVR with Hailo AI accelerator on reComputer R2000.

What you'll get:

• Real-time gun detection with bounding boxes and confidence scores
• Web dashboard for live monitoring and event review
• Automatic recording and snapshots when guns are detected
• MQTT integration for alerts and automation

Requirements: Docker on target device · Hailo AI accelerator installed

Step 1: Initialize Cameras {#init_cameras_r2000 type=manual required=false}

Set up your IP cameras and obtain RTSP stream URLs.

Wiring

1. Connect your IP camera to the same network as the deployment device
2. Find the camera's IP address (check your router's DHCP client list, or use the manufacturer's tool)
3. Access the camera's web management interface (usually http://<camera-ip>)
4. Enable RTSP streaming in the camera settings (usually enabled by default)
5. Note down the RTSP URL for your camera

Common RTSP URL formats:

• Hikvision: rtsp://admin:password@<ip>:554/Streaming/Channels/101 (main stream) or /102 (sub stream)
• Dahua: rtsp://admin:password@<ip>:554/cam/realmonitor?channel=1&subtype=0 (main) or &subtype=1 (sub)
• Generic ONVIF: Use the camera's ONVIF discovery tool to find the stream URL

Tip: Use the sub stream (lower resolution) for detection to reduce CPU/GPU load, and the main stream for recording.

You can test your RTSP URL with VLC: Media > Open Network Stream > paste the URL.

Troubleshooting

Step 2: Deploy Frigate {#deploy_frigate_r2000 type=docker_deploy required=true config=devices/r2000_hailo.yaml}

Deploy Frigate NVR with Hailo-accelerated gun detection to your reComputer R2000.

Target: Remote Deployment {#r2000_remote type=remote config=devices/r2000_hailo.yaml default=true}

Deploy to a remote reComputer R2000 via SSH.

Wiring

1. Ensure Hailo AI Kit is installed (check with ls /dev/hailo*)
2. Connect the device to the same network as your computer
3. Enter the device's IP address and SSH credentials
4. Optionally enter your RTSP camera URLs (up to 2 cameras)
5. Click Deploy to start installation

Note: The deployment will automatically check and install the required HailoRT 4.21.0 driver if needed (may take 5-10 minutes on first run). At least 4 GB of free disk space is required.

Deployment Complete

1. Open http://<device-ip>:5000 in your browser
2. You'll see two demo cameras detecting guns in sample videos
3. If you provided RTSP URLs, your cameras will also appear with gun detection enabled

To modify camera configuration later, SSH into the device and edit:

cd ~/gun-detection-frigate
nano config/config.yml
docker compose restart

Troubleshooting

Target: Local Deployment {#r2000_local type=local config=devices/r2000_hailo_local.yaml}

Deploy directly on the current machine (requires Hailo AI accelerator).

Wiring

1. Ensure Docker is installed and Hailo device is accessible (ls /dev/hailo*)
2. Click Deploy to start installation

Note: At least 4 GB of free disk space is required for the Frigate Docker image.

Deployment Complete

1. Open http://localhost:5000 in your browser
2. You'll see two demo cameras detecting guns in sample videos

To add RTSP cameras, edit the config file and restart:

nano config/config.yml
docker compose restart

Troubleshooting

Deployment Complete

Congratulations! Frigate gun detection system is now running.

Quick Verification

1. Open the Frigate dashboard in your browser (URL shown after deployment)
2. Check the Birdseye view for all-camera overview
3. Verify gun detections appear with bounding boxes on demo videos
4. Click on events to see recorded snapshots with timestamps

Adding or Modifying Cameras

To add more RTSP cameras or change existing ones, SSH into the device and edit the configuration:

cd ~/gun-detection-frigate
nano config/config.yml

Add camera entries under cameras::

  my_camera:
    enabled: true
    ffmpeg:
      inputs:
        - path: rtsp://admin:password@192.168.1.100:554/Streaming/Channels/101
          roles:
            - detect
            - record
    detect:
      width: 1920
      height: 1080
      fps: 5
    objects:
      track:
        - gun

Common RTSP URL formats:

• Hikvision: rtsp://admin:password@<ip>:554/Streaming/Channels/101
• Dahua: rtsp://admin:password@<ip>:554/cam/realmonitor?channel=1&subtype=0

Then restart Frigate:

docker compose restart

Next Steps

• Configure notification alerts via MQTT (broker runs on port 1883)
• Adjust detection thresholds in config/config.yml (objects.filters.gun.threshold)
• Set up recording retention policies (record.retain.days)
• Frigate Documentation