Abstract
Abstract. Semantic segmentation is one of the main steps in the processing chain for Airborne Laser Scanning (ALS) point clouds, but it is also one of the most labour intensive steps, as it requires many labelled examples to train a classifier. National mapping agencies (NMAs) have to acquire nationwide ALS data every couple of years for their duties. Having point clouds cover different terrains such as flat or mountainous regions, a classifier often requires a refinement using additional data from those specific terrains. In this study, we present an algorithm, which is able to classify point clouds of similar terrain types without requiring any additional training data and which is still able to achieve overall F1-Scores of over 90% in most setups. Our algorithm uses up to two height distributions within a single cell in a rasterized point cloud. For each distribution, the empirical mean and standard deviation are calculated, which are the input for a Convolutional Neural Network (CNN) classifier. Consequently, our approach only requires the geometry of point clouds, which enables also the usage of the same network structure for point clouds from other sensor systems such as Dense Image Matching. Since the mean ground level varies with the observed area, we also examined five different normalisation methods for our input in order to reduce the ground influence on the point clouds and thus increase its transferability towards other datasets. We test our trained networks on four different tests sets with the classes’ ground, building, water, non-ground and bridge.
Highlights
Semantic segmentation is the task to assign every pixel in an image or every point in a point cloud a specific label, which describes its object class
We explore and test five different height normalisation methods with varying level of complexity to remove the main part of the ground influence from the point cloud heights, while deducing what kind of complexity is necessary to achieve sufficient results using only the geometry
Instead of using the bottom mean values x, we filter the point cloud, so that only the points from set C, which have a standard deviation s lower than 0.15m, remain. 0.15m equals the absolute standard deviation from most Airborne Laser Scanning (ALS) point clouds and we discovered in our experiments, that ground points, and building points usually tend to stay within this 0.15m boundary
Summary
Semantic segmentation is the task to assign every pixel in an image or every point in a point cloud a specific label, which describes its object class. Several different approaches to generate a Digital Terrain Model (DTM) and to approximate the ground level from point clouds have been made (Chen et al, 2017). Other approaches use additional information such as aerial images, intensity values from point clouds or a mix from geometry and ALS characteristics to determine the ground level within a point cloud (Gevaert et al, 2018; Yunfei et al, 2008; Rizaldy et al, 2018). We explore and test five different height normalisation methods with varying level of complexity to remove the main part of the ground influence from the point cloud heights, while deducing what kind of complexity is necessary to achieve sufficient results using only the geometry.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
