Copyright (C) 2021, Axis Communications AB, Lund, Sweden. All Rights Reserved.
This example shows a minimal QR code detector and decoder application written in Python. It composes two different container images into an application that performs an inference using a deep learning model.
The first container contains the actual program built in this example. It then uses gRPC/protobuf to call the second container, that is used to capture images from the camera.
NumPy is used to preprocess the images and OpenCV to detect and decode any QR codes within the image.
opencv-qr-decoder-python
├── app
│ └── qr.py
├── docker-compose.yml
├── Dockerfile
└── README.md
- qr.py - The application's main script
- Dockerfile - Dockerfile specifying how the application runtime is built
- docker-compose.yml - docker-compose file specifying how/with what settings the application is started
Meet the following requirements to ensure compatibility with the example:
- Axis device
- Chip: ARTPEC-8 DLPU devices (e.g., Q1656)
- Firmware: 11.10 or higher
- Docker ACAP version 3.0 installed and started, using TLS with TCP and IPC socket and SD card as storage
- Computer
- Either Docker Desktop version 4.11.1 or higher,
- or Docker Engine version 20.10.17 or higher with BuildKit enabled using Docker Compose version 1.29.2 or higher
Define and export the application image name in APP_NAME
for use in the Docker Compose file.
export APP_NAME=acap-opencv-qr-decoder-python
# Install qemu to allow build for a different architecture
docker run --rm --privileged multiarch/qemu-user-static --credential yes --persistent yes
docker build --tag $APP_NAME .
DEVICE_IP=<actual camera IP address>
DOCKER_PORT=2376
docker --tlsverify --host tcp://$DEVICE_IP:$DOCKER_PORT system prune --all --force
If you encounter any TLS related issues, please see the TLS setup chapter regarding the DOCKER_CERT_PATH
environment variable in the Docker ACAP repository.
Browse to the application page of the Axis device:
http://<AXIS_DEVICE_IP>/index.html#apps
Click on the tab Apps
in the device GUI and enable Allow unsigned apps
toggle.
Next, the built images needs to be uploaded to the device. This can be done through a registry or directly. In this case, the direct transfer is used by piping the compressed application directly to the device's Docker client:
docker save $APP_NAME | docker --tlsverify --host tcp://$DEVICE_IP:$DOCKER_PORT load
With the application image on the device, it can be started. As the example uses OpenCV, the OpenCV requirements will be included in docker-compose.yml
, which is used to run the application:
docker --tlsverify --host tcp://$DEVICE_IP:$DOCKER_PORT compose up
# Terminate with Ctrl-C and cleanup
docker --tlsverify --host tcp://$DEVICE_IP:$DOCKER_PORT compose down --volumes