yas-sim / openvino_docker_boilerplate

A boilerplate for an OpenVINO application in a container project using an OpenVINO base container image

A boilerplate for an OpenVINO application in a container project using an OpenVINO base container image

Sample Dockerfile and OpenVINO application to demonstrate how to create a Docker project for OpenVINO using OpenVINO base container image.
The sample application uses ssd_mobilenet_v1 model and run object detection task. The application reads an image from ./input/image.jpg and write the result to ./output/image.jpg.

How to get the DL model

python prepare_model.py

How to build the image

cd Docker
docker build -t ov_app .

How to run the container

docker run --rm -v <host_working_dir>/input:/app/input -v <host_working_dir>/output:/app/output ov_app

Start the container for debugging / testing

docker run --rm -it -v <host_working_dir>/input:/app/input -v <host_working_dir>/output:/app/output ov_app /usr/bin/bash

Note:

• The application file is openvino_app.py.
• The application code and a DL model will be deployed in /app directory with the Dockerfile in this project.
• Default working directory for the container openvino/ubuntu22_runtime:2024.6.0 is /opt/intel/openvino_2024.6.0.0