There are a few solutions much better than the old tech used by meshlab. Specifically, fssr[0], which does a more advanced and especially sample-density-compens

There are a few solutions much better than the old tech used by meshlab. Specifically, fssr[0], which does a more advanced and especially sample-density-compensating variant of meshlab's poisson reconstruction, as well as instant meshes[1], which does generic retopo, and is able to handle point clouds, while still being _very_ fast and delivers surprisingly good low-poly results (see for yourself). [0]: https://www.gcc.tu-darmstadt.de/home/proj/fssr/index.en.jsp https://www.gcc.tu-darmstadt.de/home/proj/fssr/index.en.jsp [1]: https://github.com/wjakob/instant-meshes https://github.com/wjakob/instant-meshes