Procesamiento y visualización procedimental de nubes de puntos mediante software libre

Abstract

Point cloud processing in the scientific domain has become established around efficient operations for reduction, voxelization, neighborhood analysis, and filtering, typically implemented through specialized libraries external. In parallel, free and open-source software ecosystems aimed at modeling and rendering, such as Blender, provide advanced capabilities for visualization, lighting, animation, and scene composition, but they do not always include accessible workflows for the interactive preprocessing of point clouds. This paper presents a procedural tool developed in Blender using Geometry Nodes for reduction, voxelization, spatial clipping, outlier filtering, primitive instancing, and visualization of attributes associated with point clouds. The system is organized as a modular, non-destructive, and parameterizable modifier, complemented by a graphical-load estimation mechanism intended to prevent excessively costly visualization configurations. The main contribution lies in integrating basic preprocessing and graphical representation operations within a single, highly popular visual environment, thereby facilitating both preliminary data exploration and the generation of rendered images, animations, and scenes enriched with other 3D models.