Interactive Editing of Large Point Clouds

Abstract

This paper describes a new out-of-core multi-resolution data structure for real-time visualization and interactive editing of large point clouds. In addition, an editing system is discussed that makes use of the novel data structure to provide interactive editing tools for large scanner data sets. The new data structure provides efficient rendering and allows for handling very large data sets using out-of-core storage. Unlike related previous approaches, it also provides dynamic operations for online insertion, deletion and modification of points with time mostly independent of scene complexity. This permits local editing of huge models in real time while maintaining a full multi-resolution representation for visualization. The data structure is used to implement a prototypical editing system for large point clouds. It provides real-time local editing tools for huge data sets as well as a two-resolution scripting mode for planning large, non-local changes which are subsequently performed in an externally efficient offline computation. We evaluate our implementation on several synthetic and real-world examples of sizes up to 63GB.