Deploy Plate Recognizer Stream (ALPR) on ZEDEDA

​

Introduction

Plate Recognizer Stream is an AI-powered license plate recognition (ALPR) solution that runs at the edge. It detects and identifies vehicle license plates in real time, including plate number, vehicle make, model, color, and state, even in challenging conditions such as motion blur or low light. Typical use cases include transportation, logistics, smart parking, tolling, and security monitoring. This article walks through deploying stream-alpr from My Library using the New Application wizard, including persistent volume creation, network port configuration, and license injection via cloud-init.

​

Prerequisites

• Your node is onboarded to Edge Infrastructure Services and has a status of Ready under Nodes.
• The node runs a supported version of EVE-OS (Edge Virtualization Engine).
• Outbound HTTPS (port 443) connectivity is available from the node.
• You have the SysManager or SysAdmin role in your enterprise. See Manage Roles.
• A valid Plate Recognizer license key and token. Trial keys are available at platerecognizer.com/stream.

​

Create a Persistent Volume

Stream continuously processes image frames and metadata. Create a persistent volume so cached images and logs survive edge container restarts or updates.

1. Go to Infrastructure > Workloads > Volumes.
2. Click the + button in the bottom-right corner.
3. Select Block Storage as the volume type.
4. In the Identity section, enter the following values:

   Field	Value
   Name	alpr_volume_1
   Title	Enter a display title.
   Target Type	Node
   Node	Select the node you will deploy Stream to.

   Volumes are bound to a specific node at creation time. Select the same node you intend to deploy the app to in the next section.
5. Click Next. In the Storage section, enter the following values:

   Field	Value
   Access mode	Read Write
   Max size	10 GB
   Encryption	Enabled

6. Click Next to review, then click Create Volume.

​

Deploy the App

​

Add the App to Your Library

If you have not previously added the Stream app to your library:

1. Go to Marketplace > Edge Apps in the navigation.
2. Search for platerecognizer-alpr-stream and select the variant that matches your node architecture (AMD or ARM).
3. Click the + button on the card to add it to your library.

​

Start the Deploy Wizard

1. Go to Marketplace > My Library in the navigation.
2. Search for platerecognizer-alpr-stream.
3. Hover over the card and click the Deploy icon.
4. The New Application wizard opens.

​

App & Target

Click Next.

​

Interfaces

Under eth0, select the network instance to connect the app to from the dropdown. The dropdown lists available network instances on the selected node. If no suitable network instance exists, click + to create a new virtual network. Click Next.

​

Custom config

Enter your Plate Recognizer credentials in the following fields: A free trial license is available at platerecognizer.com/stream. Click Next.

​

Features

Leave the default settings or adjust as needed:

• Start Delay (seconds): Enter a value if the app should wait before starting. The default is 0.
• Enable Logs: Keep this enabled to view app logs in Edge Infrastructure Services.

Click Next.

​

Review

Review your configuration and click Deploy.

​

Verify the Deployment

1. Go to Applications and confirm the run state is Online.
2. Open the Logs tab and confirm the following startup message appears:
   INFO:root:Plate Recognizer Stream v1.59.0 started successfully
   
3. Test connectivity using your RTSP camera stream or by accessing http://<node-ip>:8001/status/.

​

Hardware Sizing

The number of concurrent camera feeds that perform well depends on the CPU, GPU, and memory available on your node. For detailed system requirements, see the Plate Recognizer hardware sizing guide.

​

Next Steps