How to Set Up an LPR Camera with RTSP and IP Streams

set up lpr rtsp ip camera

Any IP camera that streams via RTSP can be connected to license plate recognition software in four steps: confirm the RTSP stream works, enter the stream URL into the LPR software dashboard, configure the capture zone, and test with a real vehicle. The entire process takes 15–45 minutes for a single camera, including troubleshooting time.

Before You Start: Prerequisites Checklist

Verify these items before opening the LPR software dashboard:

• Camera resolution: 720p minimum (1080p or higher strongly recommended)
• Camera is powered and connected to your local network via Ethernet
• You know the camera’s local IP address (check your router’s device list)
• You have the camera’s admin username and password
• The network has internet access (required for cloud-based LPR like placa.ai)
• Camera is aimed at the vehicle lane — license plate visible for at least 1–2 seconds as vehicles pass

Step 1: Find Your Camera’s RTSP Stream URL

Every IP camera that supports RTSP has a specific URL format for accessing its video stream. Find your camera’s brand below:

Replace [user], [pass], and [ip] with your camera’s admin username, password, and local IP address (e.g., 192.168.1.100).

Step 2: Test the Stream in VLC Before Connecting to LPR Software

Always verify the RTSP stream works before entering it into your LPR software. This confirms the URL is correct and eliminates the camera as a cause of any later issues.

1. Download and open VLC Media Player (free, available at videolan.org)
2. Go to Media → Open Network Stream (or press Ctrl+N)
3. Paste your RTSP URL into the field
4. Click Play

If you see live video, the stream is working. If VLC shows a black screen or error, see the troubleshooting section below.

Step 3: Connect the Camera to placa.ai

1. Log into your placa.ai dashboard
2. Click Add Camera
3. Enter your camera’s RTSP URL in the stream field
4. Give the camera a name (e.g., “Front Gate” or “Lot B Entrance”)
5. Click Connect — the system will verify the stream and display a live preview
6. If the preview loads, the camera is successfully connected

Connection typically takes 10–30 seconds. If it times out, verify the camera and the computer running the placa.ai agent are on the same local network.

Step 4: Configure the Capture Zone

A capture zone is the region of the camera frame where placa.ai looks for license plates. Configuring a precise zone improves accuracy and reduces false reads from vehicles on adjacent lanes or streets.

1. In the camera settings panel, click Set Capture Zone
2. Draw a polygon around the area where vehicles will have their plates visible — typically the width of one lane at the point where the vehicle slows or stops
3. Aim for a zone that covers 2–3 car lengths — enough time to capture a clear plate read as the vehicle approaches
4. Avoid including the road edge or adjacent lanes in the zone if possible
5. Save the zone and enable plate detection

Step 5: Configure Your Plates Database

For gate access control, set up your approved plates list (whitelist):

1. Go to Plates → Manage Lists
2. Create a whitelist named after your community or property (e.g., “Resident Vehicles”)
3. Add plates manually or upload a CSV with plate numbers, owner names, and expiration dates
4. Assign the whitelist to your camera’s gate action — when a whitelisted plate is read, the gate opens signal fires
5. Set up an alert rule for unrecognized plates: email, SMS, or in-app notification

Step 6: Test with a Real Vehicle

Before going live, drive a test vehicle through the capture zone:

1. Check the Recent Reads log in the dashboard — you should see the plate appear within 2 seconds of passing the camera
2. Verify the read is accurate (compare the logged plate to the actual plate on the vehicle)
3. If using gate integration, verify the gate opens for whitelisted plates and stays closed for unrecognized plates
4. Test at night with headlights on to verify IR illumination and nighttime accuracy

Camera Placement Tips for Maximum Accuracy

Angle

Mount the camera at a 20–40° downward angle, aimed at the license plate level as vehicles approach. Too steep (looking straight down) makes plates hard to read. Too shallow (camera at car height) makes the plate appear at a severe angle.

Distance

Optimal capture distance is 10–25 feet from the camera to the vehicle. At this range, a 1080p camera can resolve individual plate characters clearly. Beyond 30 feet, 4MP or higher resolution is recommended.

Horizontal position

Mount the camera slightly off-center from the lane — about 5–10 feet to the side. This reduces direct IR reflection off the plate at night (the retroreflective surface bounces light straight back at the source).

Lighting

For 24/7 operation, use a camera with IR night vision. If the camera is inside a structure (parking garage, covered gate), ambient light may be sufficient. Avoid pointing the camera directly into a light source — backlighting washes out the plate.

Troubleshooting Common Setup Errors

IR Camera Configuration for Night Accuracy

Night performance is the most common setup challenge. License plates are retroreflective by design — they bounce IR light directly back at the camera, causing the plate to appear as a bright white rectangle with no readable characters.

Settings to adjust in your camera’s web interface:

1. Log into the camera’s web interface (type the camera’s IP address in a browser)
2. Find Image Settings → IR/Light Control
3. Reduce IR intensity from 100% to 40–60% for close gates (under 20 feet)
4. Enable WDR (Wide Dynamic Range) if available — this helps balance bright plates against dark backgrounds
5. Set Day/Night Mode to Auto so the camera switches to IR mode automatically after dark
6. Re-test with a vehicle at night after making adjustments

Frequently Asked Questions

Does placa.ai require a dedicated server or PC to run?

For cloud-based processing, placa.ai requires only a lightweight local agent (a small software process that runs on any Windows or Linux PC or NAS device on your network). This agent connects cameras to the cloud processing engine. No dedicated server hardware is required.

Can I connect cameras from different brands to the same placa.ai account?

Yes. Since placa.ai connects via RTSP streams, cameras from different manufacturers can all feed into the same dashboard. Each camera gets its own stream URL and capture zone configuration. A single account can manage a Hikvision camera at the front gate and a Reolink camera at the parking lot entry simultaneously.

How do I find my camera’s IP address?

Three methods: (1) Check your router’s admin interface under connected devices; (2) Use the camera manufacturer’s IP finder tool (Hikvision SADP, Dahua Device Manager, Reolink app); (3) Check the camera’s sticker — some units have the default IP printed on the label.

My camera has ONVIF but I can’t find the RTSP URL. What do I do?

ONVIF-compatible cameras expose their RTSP URL via the ONVIF protocol. Download the free ONVIF Device Manager tool, run a scan on your network, and it will discover the camera and display its RTSP stream URL. This works for any ONVIF-compliant camera regardless of brand.

What frame rate do I need for moving vehicles?

For gate applications where vehicles slow to 5–10 mph, 15 fps is the minimum and 25–30 fps is preferred. At speeds above 15 mph (parking lot through-lanes), frame rate becomes more critical — use 30 fps at minimum and position the camera for a longer approach distance to give the software more frames to work with.

Reference: Automatic Number Plate Recognition – Wikipedia