Utilizing Machine Learning for Training and Custom Automatic Point Cloud Classification

Keywords: Machine Learning, Custom Point Cloud Classification, UAV, Drone, Global Mapper Pro, Automatic Classification, Point Cloud Analysis
Abstract Type: Paper Abstract

Abstract

The increased availability of UAVs/drones capable of collecting high-resolution point clouds introduces the need to identify and measure more features within those point clouds. Automatic classification methods for identifying standard features such as ground and vegetation are commonly available, but not many custom classification tools exist outside of costly development. Unique or industry-specific feature classifications require expensive solutions of tailored software or person-hours spent in manual point cloud classification. Applying machine learning techniques to identify unique features expands what is possible when it comes to point cloud classification. These techniques provide the ability to train and create custom classifications that can automatically identify target features within a point cloud. This functionality is included in Global Mapper Pro and increases the accessibility of cutting-edge point cloud analysis in any industry without a high-cost barrier.