Abstract
Classification is a fundamental task for airborne laser scanning (ALS) point cloud processing and applications. This task is challenging due to outdoor scenes with high complexity and point clouds with irregular distribution. Many existing methods based on deep learning techniques have drawbacks, such as complex pre/post-processing steps, an expensive sampling cost, and a limited receptive field size. In this paper, we propose a graph attention feature fusion network (GAFFNet) that can achieve a satisfactory classification performance by capturing wider contextual information of the ALS point cloud. Based on the graph attention mechanism, we first design a neighborhood feature fusion unit and an extended neighborhood feature fusion block, which effectively increases the receptive field for each point. On this basis, we further design a neural network based on encoder–decoder architecture to obtain the semantic features of point clouds at different levels, allowing us to achieve a more accurate classification. We evaluate the performance of our method on a publicly available ALS point cloud dataset provided by the International Society for Photogrammetry and Remote Sensing (ISPRS). The experimental results show that our method can effectively distinguish nine types of ground objects. We achieve more satisfactory results on different evaluation metrics when compared with the results obtained via other approaches.
Highlights
Airborne laser scanning (ALS), known as airborne light detection and ranging (LiDAR), is an important active remote sensing technique that has displayed rapid development in recent years [1]
Researchers classified point clouds by employing hand-engineered features and traditional classifiers [9,10,11,12] or preprocessed the point clouds before classification [13,14]. These methods belong to traditional machine learning methods, which fail to learn highlevel features, whereas the methods based on deep learning techniques can further improve the classification accuracy due to the ability to learn high-level features
Many previous works on point cloud classification based on the graph attention mechanism have achieved good results, but few of them have solved the problem that the receptive field size of the networks is limited
Summary
Airborne laser scanning (ALS), known as airborne light detection and ranging (LiDAR), is an important active remote sensing technique that has displayed rapid development in recent years [1]. Transform 3D point clouds into 2D images or 3D grids and use deep learning techniques for classification; the transformation leads to information loss and a high computation cost To avoid these problems, some studies directly process raw points while employing deep learning techniques, such as PointNet++ [17], SPG [18], and RandLANet [19]. Some studies directly process raw points while employing deep learning techniques, such as PointNet++ [17], SPG [18], and RandLANet [19] The latter two networks have achieved good results in the large-scale point cloud classification task, which is considerably challenging. Many works capture more local features of the point cloud data by introducing a graph neural network [20] and a graph attention mechanism [21].
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
