Image Segmentation
This tool performs image segmentation, following the technique by Baatz and Schäpe (2000).
It has the following parameters:
- Compactness weight.
- Colour weight.
- Maximum allowable cost function.
- Use DBSCAN to group segments.
- DBSCAN eps - The maximum distance between the mean signature between two segments, for one to be considered as in the neighborhood of the other. The input data is scaled using a standard scaler function (removing the mean and scaling to unit variance).
References
Baatz, M, Schäpe, A., 2000. Multiresolution segmentation: An optimization approach for high quality multi-scale image segmentation. Proceedings of Angewandte Geographische Informationsverarbeitung XII, 12-23