Recent Advancements in Learning Algorithms for Point Clouds: An Updated Overview

The latest 3D acquisition mechanisms have enabled the modeling of real 3D scenes by means of unordered sets of 3D points, which can be accompanied by different attributes, e.g., color components, normals, semantic labels, and sensing-related measurements. Point clouds require an increment of the storage space, as well as of the processing computational load with respect to bi-dimensional images. Different data structures have been proposed in order to enable efficient handling of the acquired data. Data and algorithm selections are strongly driven by the requirements of specific applications. As a result, it is possible to distinguish different types of point cloud data, depending on the technologies used for the acquisition or generation.