Why does Avigilon LPR Performance Mode disable Face Recognition on AI NVRs?
Answered by ANPR AI·Sourced
Short answer
Activating high-performance LPR heavily taxes the AI NVR's computational resources. To prevent frame dropping and maintain accurate plate reads on fast-moving vehicles, the system redirects all available processing power to the OCR engine, intentionally disabling conflicting analytics like Face Recognition.
Explanation
Modern AI NVRs utilize System-on-Chip (SoC) architectures with a finite pool of tensor processing power for running deep learning models. License Plate Recognition is inherently one of the most resource-intensive analytics available, as it requires analyzing high-framerate, high-resolution video to capture vehicles traveling at speed. When administrators enable "LPR Performance Mode" on an Avigilon AI NVR, the software must prioritize that channel's stream.
To ensure the LPR engine has the necessary compute budget, Avigilon's software intentionally disables other intelligent analytics that would normally share that processor. Depending on the hardware tier (Premium, Standard, or Value), features such as Face Recognition, Face Mask Detection, and Classified Object Detection will completely stop functioning across the network while LPR Performance Mode is active.
Things to check
- Identify which tier of Avigilon AI NVR is deployed (Premium, Standard, or Value), as this dictates exactly how many analytics features are disabled.
- Check the "Max Image Analysis Rate" configuration; setting this too high will consume massive processing capacity.
What to do
You must carefully calculate the compute budget of your hardware. If you require both perimeter Face Recognition and LPR simultaneously, you cannot run them on the same low-tier NVR. Depending on the model, a Premium AI NVR can run up to three regional license plate configurations concurrently while disabling face mask detection.
If all analytics are mission-critical, you must separate the LPR processing onto dedicated hardware channels or utilize edge-capable LPR cameras that perform the OCR internally before sending the data to the NVR.
ANPR Edge
Skip the dedicated ANPR camera. Read plates from the cameras you already have.
ANPR Edge is plate-recognition software that runs on a PC at your site and works with ordinary IP cameras, so plate reads no longer depend on one vendor's camera, NVR or firmware.
- Works with standard IP cameras over RTSP
- One license per location covers every camera
- Runs on a Windows PC or Mac you own