Eike Barnefske

Point clouds are generated by light imaging, detection and ranging (LIDAR) scanners or depth imaging cameras, which capture the geometry from the scanned objects with high accuracy. Unfortunately these systems are unable to identify the semantics of the objects. Semantic 3D point clouds are an important basis for modeling the real world in digital applications. Manual semantic segmentation is a labor and cost intensive task. Automation of semantic segmentation using machine learning and deep learning (DL) approaches is therefore an interesting subject of research. In particular, point-based network architectures, such as PointNet, lead to a beneficial semantic segmentation in individual applications. For the application of DL methods, a large number of hyperparameters (HP) have to be determined and these HPs influence the training success. In our work, the investigated HP are the class distribution and the class combination. By means of seven combinations of classes following a hierarchical scheme and four methods to adapt the class sizes, these HPs are investigated in a detailed and structured manner. The investigated settings show an increased semantic segmentation performance, by an increase of 31% in recall for the class Erroneous points or that all classes have a recall of higher than 50%. However, based on our results the correct setting of only these HP does not lead to a simple, universal and practical semantic segmentation procedure.

Recently, 3D point clouds have become a quasi-standard for digitization. Point cloud processing remains a challenge due to the complex and unstructured nature of point clouds. Currently, most automatic point cloud segmentation methods are data-based and gain knowledge from manually segmented ground truth (GT) point clouds. The creation of GT point clouds by capturing data with an optical sensor and then performing a manual or semi-automatic segmentation is a less studied research field. Usually, GT point clouds are semantically segmented only once and considered to be free of semantic errors. In this work, it is shown that this assumption has no overall validity if the reality is to be represented by a semantic point cloud. Our quality model has been developed to describe and evaluate semantic GT point clouds and their manual creation processes. It is applied on our dataset and publicly available point cloud datasets. Furthermore, we believe that this quality model contributes to the objective evaluation and comparability of data-based segmentation algorithms.

Point clouds give a very detailed and sometimes very accurate representation of the geometry of captured objects. In surveying, point clouds captured with laser scanners or camera systems are an intermediate result that must be processed further. Often the point cloud has to be divided into regions of similar types (object classes) for the next process steps. These classifications are very time-consuming and cost-intensive compared to acquisition. In order to automate this process step, conventional neural networks (ConvNet), which take over the classification task, are investigated in detail. In addition to the network architecture, the classification performance of a ConvNet depends on the training data with which the task is learned. This paper presents and evaluates the point clould classification tool (PCCT) developed at HCU Hamburg. With the PCCT, large point cloud collections can be semi-automatically classified. Furthermore, the influence of erroneous points in three-dimensional point clouds is investigated. The network architecture PointNet is used for this investigation.