This implements an edge detection processing block in Edge Impulse. How to use:
-
Docker:
-
Build the container:
$ docker build -t edge-detection-block .
-
Run the block:
$ docker run -p 4446:4446 -it --rm edge-detection-block
-
-
Or, locally:
-
Install Python 3.8.
-
Install the dependencies:
$ pip3 install -r requirements-blocks.txt
-
Run the block:
$ python3 dsp-server.py
-
-
Install ngrok and open up port 4446 to the world:
$ ngrok http 4446
Note down the 'Forwarding' address that starts with https, e.g.:
Forwarding https://4e9e1e61e3aa.ngrok.io -> http://localhost:4446
-
In Edge Impulse, go to Create impulse, click Add a processing block, click Add custom block and enter the URL from the previous step.
-
You now have edge detection as a preprocessing block:
-
Train your model as usual ๐