besn0847 / arm-yolo

A microservice for Raspberry PI running a Yolo neural network to detect object in pictures

arm-yolo

A microservice for Raspberry PI running a Yolo neural network to detect object in pictures

To start the container with default config :

docker run -d --name yolo -p 5000:5000 x86-yolo

Alternatively you can specify a conf diretory so you can change model weights easily :

docker run -d --name yolo -v <conf_dir>:/conf -p 5000:5000 x86-yolo

To test an image with the Yolo WebService :

curl -X PUT -F image_file=@/home/franck/Sandbox/x86-yolo/image.jpg http://localhost:5000/process

Result is something like : Class: cat with confidence: 0.9518834352493286

Based on Alpine Linux 3.8, Python 3.6 and OpenCV 3.4.4.

Tributes

1. https://github.com/arunponnusamy/object-detection-opencv
2. https://hub.docker.com/r/julianbei/alpine-opencv-microimage/