Integrate a Python Algorithm

This guide will show you how to setup your workspace to integrate your own Python motion segmentation algorithm using the Python PLI. It is assumed that you either followed the quickstart guide or you compiled the MSeg Core Module from source and set up the Python PLI.

Install the Skeleton Project

We will use the mseg generate terminal tool to install a skeleton project.

# Define the base directory where the project directory should be located
$ export $PROJECT_BASE_DIR=~/projects
# Run the generation tool to create the skeleton.
# If your algorithm requires training, add the --train flag
$ mseg generate --python $PROJECT_BASE_DIR ExampleAlgorithm

Run the Algorithm

To run the algorithm, both ArmarX and the MSeg core module must be running. You can then start the algorithm like this:

# Change into the project folder
$ cd $PROJECT_BASE_DIR/examplealgorithm
# Run the algorithm
$ python

You should now see an output like this:

user@machine:~/projects/examplealgorithm$ python
Connecting to MSeg core module...
Connection established
ExampleAlgorithm ready

You can stop the algorithm with Ctrl + C at any time.

How to Proceed

You can now start implementing your algorithm. You may also want to get more familiar with the SegmentationAlgorithm API reference, the MSeg GUI and the mseg terminal tool.