An update on Random3Dcity

It has already been two months since I have released Random3Dcity, an experimental basic procedural modelling engine that natively supports CityGML. It is an experimental research prototype that I have built for my PhD project, but nevertheless it has generated quite some interest in the 3D GIS community.

For instance, it has been featured on the ISPRS datasets webpage and in the ISPRS newsletter of December 2014. Nice promotion.

Another thing worth mentioning is that my friend Marko Čubranić found it interesting and he has decided to 3D print a sample dataset. I guess that goes to history books as the first 3D printed CityGML procedurally generated model 🙂

3D printed CityGML dataset with Ultimaker II. Sample of Random3Dcity.

3D printed CityGML dataset with Ultimaker II. Sample of 144 gridded buildings with a street network generated with Random3Dcity.

The most important news is that I have updated the code so now it works with Python 3 (thanks to Mickaël Brasebin for the suggestions). I have also squished a bug where an lod3MultiSurface could have been composed of multiple gml:MultiSurfaces, which is a violation of the CityGML 2.0 standard.

Bug reports and suggestions are always welcome.