I’m trying to do the following:

1. I’ve got a working yolov8 custom model trained and segmenting intersections as expected.
   It predicts objects like this:

2. What I want to do is to convert the segmented items above into polygons or lines, ie:

To clarify, the image above (2) would be a outcome of a geojson file, ie:

"type": "FeatureCollection",
"features": [
    {
        "type": "Feature",
        "geometry": {
            "type": "LineString",
            "coordinates": [
                [
                    9689306.142846698,
                    4424394.307639465
                ],
                [
                    9689296.790935714,
                    4424382.4802226419
                ]
            ]
        },
        "properties": {
            "pedId": 2,
            "domain": "pedestrian_crossing",
            "width": 3.0,
            "nearestRoadId": 2
        }
    },

Does anyone know a way to make this conversion? (results.masks -> GeoJSON)