Name	Name	Last commit message	Last commit date
parent directory ..
Dockerfile	Dockerfile
Dockerfile.cpu	Dockerfile.cpu
README.md	README.md

MareArts ANPR Server — Docker

Run the MareArts ANPR server as a Docker container — with GPU (CUDA) or CPU-only.

The Docker image includes the complete ANPR server with 20+ REST API endpoints, web dashboard, Vehicle Intelligence, and all V16 models — ready to deploy in seconds.

Dashboard	Detection	History

Two Images

	GPU (CUDA)	CPU
Dockerfile	`Dockerfile`	`Dockerfile.cpu`
Base	`nvidia/cuda:12.5.1-cudnn-runtime-ubuntu22.04`	`ubuntu:22.04`
Python	3.13	3.13
Runtime	`onnxruntime-gpu` (CUDA + TensorRT + CPU fallback)	`onnxruntime` (CPU only)
Image size	~5-6 GB	~1-1.5 GB
Requires	NVIDIA GPU + Container Toolkit	Any machine with Docker
Best for	Production, high throughput	Development, testing, no-GPU servers

Prerequisites

Docker 20.10+
NVIDIA Container Toolkit (GPU image only) — Install Guide
MareArts License — Get one here

Quick Start — GPU

docker build -t marearts-anpr-server:latest .

docker run -d \
  --gpus all \
  --name marearts-anpr-server \
  -p 8000:8000 \
  -e MAREARTS_ANPR_USERNAME="your@email.com" \
  -e MAREARTS_ANPR_SERIAL_KEY="your_serial_key" \
  -e MAREARTS_ANPR_SIGNATURE="your_signature" \
  -v ~/.marearts:/root/.marearts \
  marearts-anpr-server:latest

Quick Start — CPU

docker build -t marearts-anpr-server-cpu:latest -f Dockerfile.cpu .

docker run -d \
  --name marearts-anpr-server-cpu \
  -p 8000:8000 \
  -e MAREARTS_ANPR_USERNAME="your@email.com" \
  -e MAREARTS_ANPR_SERIAL_KEY="your_serial_key" \
  -e MAREARTS_ANPR_SIGNATURE="your_signature" \
  -v ~/.marearts:/root/.marearts \
  marearts-anpr-server-cpu:latest

No --gpus flag needed for CPU. Works on any machine — Linux, macOS, Windows, ARM, cloud VMs without GPU.

Use it

# Health check
curl http://localhost:8000/api/health

# Detect plates
curl -X POST http://localhost:8000/api/anpr -F "image=@car.jpg"

# Vehicle Intelligence (make, model, color, type)
curl -X POST http://localhost:8000/api/anpr/mmc -F "image=@car.jpg"

# Open dashboard in browser
open http://localhost:8000

Build from Source

git clone https://github.com/MareArts/MareArts-ANPR.git
cd MareArts-ANPR/docker

# GPU image
docker build -t marearts-anpr-server:latest .

# CPU image
docker build -t marearts-anpr-server-cpu:latest -f Dockerfile.cpu .

Credentials

The server needs your MareArts license credentials. Three options:

Option 1: Environment Variables (recommended)

docker run -d \
  -e MAREARTS_ANPR_USERNAME="your@email.com" \
  -e MAREARTS_ANPR_SERIAL_KEY="your_serial_key" \
  -e MAREARTS_ANPR_SIGNATURE="your_signature" \
  ...

Option 2: Credentials File

Create ~/.marearts/.marearts_env:

export MAREARTS_ANPR_USERNAME="your@email.com"
export MAREARTS_ANPR_SERIAL_KEY="your_serial_key"
export MAREARTS_ANPR_SIGNATURE="your_signature"

Then mount the directory:

docker run -d \
  -v ~/.marearts:/root/.marearts \
  ...

Option 3: Interactive Setup

docker run -it --rm \
  -v ~/.marearts:/root/.marearts \
  marearts-anpr-server:latest \
  ma-anpr config

This saves credentials to ~/.marearts/ for future use.

Environment Variables

Variable	Purpose	Default
`MAREARTS_ANPR_USERNAME`	License email	(required)
`MAREARTS_ANPR_SERIAL_KEY`	License serial key	(required)
`MAREARTS_ANPR_SIGNATURE`	License signature	(required)
`ANPR_BACKEND`	Inference backend: `auto`, `cuda`, `cpu`	`auto` (GPU) / `cpu` (CPU image)
`ANPR_HOST`	Server bind address	`0.0.0.0`
`ANPR_PORT`	Server port	`8000`
`ANPR_REGION`	Default OCR region	`univ`
`ANPR_THREADS`	Worker threads	`4`

API Endpoints

Once the server is running, you have access to 20+ endpoints:

Endpoint	Method	Description
`/api/health`	GET	Server health and status
`/api/anpr`	POST	Detect and read plates
`/api/anpr/mmc`	POST	Vehicle Intelligence (make, model, color, type)
`/api/config`	GET	Server configuration
`/api/stats`	GET	Request statistics
`/api/history`	GET	Detection history
`/api/watchlist`	GET/POST	Manage plate watchlist
`/api/alerts`	GET	View triggered alerts
`/api/regions`	GET	Available OCR regions
`/api/models`	GET	Loaded model info
`/docs`	GET	Interactive API documentation (Swagger)

Full API documentation is available at http://localhost:8000/docs when the server is running.

Docker Compose

GPU

version: "3.8"
services:
  anpr-server:
    image: marearts-anpr-server:latest
    ports:
      - "8000:8000"
    environment:
      - MAREARTS_ANPR_USERNAME=your@email.com
      - MAREARTS_ANPR_SERIAL_KEY=your_serial_key
      - MAREARTS_ANPR_SIGNATURE=your_signature
      - ANPR_BACKEND=auto
    volumes:
      - ~/.marearts:/root/.marearts
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8000/api/health"]
      interval: 60s
      timeout: 30s
      start_period: 300s
      retries: 5

CPU

version: "3.8"
services:
  anpr-server:
    image: marearts-anpr-server-cpu:latest
    ports:
      - "8000:8000"
    environment:
      - MAREARTS_ANPR_USERNAME=your@email.com
      - MAREARTS_ANPR_SERIAL_KEY=your_serial_key
      - MAREARTS_ANPR_SIGNATURE=your_signature
      - ANPR_BACKEND=cpu
    volumes:
      - ~/.marearts:/root/.marearts
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8000/api/health"]
      interval: 60s
      timeout: 30s
      start_period: 300s
      retries: 5

docker compose up -d

Useful Commands

# View logs
docker logs -f marearts-anpr-server

# Check health
docker exec marearts-anpr-server curl -s http://localhost:8000/api/health

# Verify providers (GPU image)
docker exec marearts-anpr-server python -c "import onnxruntime; print(onnxruntime.get_available_providers())"
# Expected: ['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'CPUExecutionProvider']

# Verify providers (CPU image)
docker exec marearts-anpr-server-cpu python -c "import onnxruntime; print(onnxruntime.get_available_providers())"
# Expected: ['CPUExecutionProvider']

# Stop the server
docker stop marearts-anpr-server

# Remove the container
docker rm marearts-anpr-server

Support


Homepage	marearts.com
License	marearts.com/products/anpr
Live Demo	live.marearts.com
Contact	hello@marearts.com
Server Docs	Server README
SDK Docs	Python SDK README

FilesExpand file tree

docker

Directory actions

More options