and a touch of typical AI overview...
To get individual GPS coordinates for every individual parking stall, you must switch to a different pre-trained model or add a post-processing tool.
1. Use the "Parking Spot Detection" Model Instead
- What it does: It uses a MaskRCNN architecture trained specifically to outline individual vacant and occupied parking spaces.
- Input needed: Ultra-high-resolution aerial imagery (typically 7–30 cm spatial resolution). [1, 2]
2. The Step-by-Step Geoprocessing Workflow
Once you run the correct Parking Spot Detection model, you will have thousands of tiny individual polygons on your map. To turn those into a list of GPS coordinates, use this sequence in ArcGIS Pro:
- Convert Polygons to Points: Open the Feature to Point tool. Input your detected parking spots layer. This will place a single point directly in the center (centroid) of every single parking stall.
- Calculate the GPS Coordinates: Open the Add Geometry Attributes tool. Select your new points layer, choose COORDINATE_SYSTEM as your geometry property, and set your coordinate system to WGS 1984 (which outputs standard GPS latitude and longitude).
- Export to Spreadsheet: Right-click the attribute table of your points layer and select Export Table. Save it as a .csv file. You will now have a clean spreadsheet containing the unique Latitude and Longitude for every parking space.
If you want to try this route, do you already have ArcGIS Pro with the Deep Learning Libraries installed
