pullmyleg/darknet repository overview

⁠Simple AlexyAB/Darknet CPU docker Container

Raw ubuntu image with OpenCV and the latest AlexeyAB/darknet installed. Also includes yolov4-tiny weights and a horse testing video (takes a little while to find objects).

⁠To use:

1. Install docker

2. run: docker pull pullmyleg/darknet

3. run: docker run -p 8070:8070 -p 8090:8090 --name darknet_pullmyleg -it pullmyleg/darknet

4. Within the container run: $ ./darknet detector demo cfg/coco.data cfg/yolov4-tiny.cfg yolov4-tiny.weights horse.mp4 -json_port 8070 -mjpeg_port 8090 -ext_output -dont_show

5. Open: http://localhost:8070/⁠ on your local machine and see the detection output. Open http://localhost:8090/⁠ to see the video stream.

6. For more information on commands that can be run within this container see: https://github.com/AlexeyAB/darknet#how-to-use-on-the-command-line⁠

⁠To build yourself from dockerfile:

This option allows you to install more requirements and to modify the Makefile, e.g. use CUDA and GPU.

1. Install docker

2. Save dockerfile as dockerfile on local machine

3. Navigate to this location in powershell, cmd, terminal etc.

4. run: docker build -t darknet .

5. run: to the fridge for a beer (openCV is slow)

6. Run: docker run -p 8070:8070 -p 8090:8090 --name darknet -it darknet

7. Within the container run: $ ./darknet detector demo cfg/coco.data cfg/yolov4-tiny.cfg yolov4-tiny.weights horse.mp4 -json_port 8070 -mjpeg_port 8090 -ext_output -dont_show

8. Open: http://localhost:8070/⁠ on your local machine and see the detection output. Open http://localhost:8090/⁠ to see the video stream.

9. For more information on commands that can be run within this container see: https://github.com/AlexeyAB/darknet#how-to-use-on-the-command-line⁠