Formats and Compatibility
What file types are supported?
Full support for the two most common formats in the industry: LAS, LAZ (compressed), with the option of obtaining them directly from URLs.
Full support for the two most common formats in the industry: LAS, LAZ (compressed), with the option of obtaining them directly from URLs.
The framework is designed for deployment on HPC (High Performance Computing) clusters. For preprocessing and classic Machine Learning tasks, VL3D++ runs on conventional workstations. However, for Deep Learning runs, high-end equipment with last-generation GPUs of at least 32 GB VRAM is recommended to ensure optimal performance on large-scale clouds.
No. The framework offers preconfigured mining, training and prediction pipelines for the most common cases, allowing professional results to be obtained with a minimal configuration in Python.
Yes. Thanks to its tile-based processing architecture and intelligent decimation, it is capable of analysing entire territorial extents without RAM limitations.
Yes, the framework is distributed under the MIT Licence. This is a highly permissive open-source licence that allows use, copy, modification, merger, publication, distribution and sale of the software, both for academic and commercial purposes, with the sole condition of including the original copyright notice.
