Advanced On-Premise NVR System Development

Job ID: 39557229

Budget: $1,500 – $3,000 USD

DO NOT POST RANDOM PRICE HOLDERS. Submit an accurate price. Feel free to message me any questions.

I am looking for an Enterprise-Grade NVR System with LPR, Face Recognition, and Optional Vehicle Classification (Frigate+, YOLO, or Alternative)

We are building a secure, on-premise NVR system for enterprise use. Our top priorities are:

License Plate Recognition (LPR)

Face Detection and/or Recognition

Real-time or near real-time performance

On-premise processing only (no cloud dependencies)

Scalable across multiple cameras and locations

Efficient resource usage (Coral TPU, GPU, or similar)

Optional capabilities we’re interested in:

General vehicle type detection (e.g. sedan, SUV, truck)

Fine-grained vehicle identification, specifically distinguishing between models such as Toyota Hilux vs Nissan Navara

Vehicle color detection

We are currently considering Frigate+ with YOLO, but are open to better or more efficient alternatives that can meet these goals.

It would need an NVR interface for live and playback, but that will depend on the base we use, freight, shinobi, etc


Requirements:

Proven experience building or deploying video analytics systems at scale

Familiarity with detection pipelines and integration of LPR and facial recognition

Ability to trigger OCR or facial match pipelines from detection events

Capable of optimizing system performance for low-latency detection

Must deliver working configurations/scripts and clear documentation

Nice to have:

Experience with Frigate+ and custom YOLO training

Integration with Plate Recognizer, CompreFace, or similar

Setup of alerts, search tools, or dashboards for reviewed footage

To apply:

Describe your relevant experience with similar projects

Outline your recommended tech stack and why

Share links to your work (GitHub, screenshots, videos, or brief case studies)