A Troubleshooting Guide for Lorex Facial Recognition Issues
Lorex Facial Recognition is a powerful smart feature that elevates your security system from simple motion detection to intelligent person identification. When working correctly, it can tell you not just that someone is at your door, but whether it's a recognised family member or a stranger. However, when it fails to detect faces or struggles with recognition, it can be a significant source of frustration.
If you're getting poor results from your Lorex system, don't worry. This guide will help you troubleshoot and optimise your setup for accurate and reliable facial recognition.
How Facial Recognition Should Work
- Detection: The camera first identifies a human face in its field of view.
- Analysis: It captures the facial features and creates a digital template.
- Comparison: It compares this template against a database of faces you have registered.
- Alert: It sends you a specific notification, such as "Familiar face: John" or "Unfamiliar face detected."
Fixing Common Lorex Facial Recognition Problems
Let's walk through the key areas that impact the performance of this smart feature.
1. Enable the Correct Smart Plan
For many Lorex NVRs, you must tell the system which smart feature to use for each camera channel. You can't run Face Detection and, for example, Vehicle Detection on the same camera simultaneously.
- Log into your Lorex NVR using a connected mouse and monitor.
- Go to Main Menu > Events.
- In the left-hand menu, select Smart Plan.
- Choose the camera channel you want to use for facial recognition.
- Click the 'Face Detection' icon to make it the active plan for that camera.
- Click Apply to save the changes.
2. Optimise Camera Placement and Angle
This is the most critical factor for accurate detection. If the camera can't get a clear view of a person's face, it cannot work.
- Height: Mount the camera between 7 and 10 feet from the ground. Any higher, and you will only see the top of people's heads.
- Angle: The camera should be angled down slightly, not pointing straight out. The ideal angle captures faces as people walk towards the camera.
- Distance: Facial recognition is most effective when subjects are within 5 to 20 feet of the camera.
- Lighting: Ensure the area is well-lit. Avoid pointing the camera at a bright light source (backlighting), which will create silhouettes. At night, ensure the camera's IR night vision provides clear illumination of faces.
3. Configure Face Detection Settings
Once the smart plan is active, you need to fine-tune the detection settings.
- On your NVR, go to Events > Event Settings > Face Detection.
- Enable the feature for the desired channel.
- Draw a Detection Area: Create a zone where you want the camera to look for faces. This should be an area where you expect to get a clear, straight-on view of a person, like a pathway or doorway.
- Adjust Sensitivity and Thresholds: You can often adjust sensitivity settings. A higher sensitivity will detect faces more easily but may lead to more false positives. Start with the default and adjust as needed based on performance.
4. Build a High-Quality Face Database
The system's ability to recognise familiar people depends entirely on the quality of the images you provide.
- Navigate to the Face Database management section on your NVR.
- When registering a new person, use multiple high-quality images.
- Use clear, well-lit photos where the person is looking directly at the camera.
- Avoid pictures with hats, sunglasses, or shadows over the face.
- Add several images for each person from slightly different angles to improve the chances of a successful match.
By systematically checking and optimising these four key areas—Smart Plan, camera placement, detection settings, and your face database—you can significantly improve the accuracy and reliability of your Lorex facial recognition system.