Abstract
The classification and recognition of the shapes of buildings in map space play an important role in spatial cognition, cartographic generalization, and map updating. As buildings in map space are often represented as the vector data, research was conducted to learn the feature representations of the buildings and recognize their shapes based on graph neural networks. Due to the principles of graph neural networks, it is necessary to construct a graph to represent the adjacency relationships between the points (i.e., the vertices of the polygons shaping the buildings), and extract a list of geometric features for each point. This paper proposes a deep point convolutional network to recognize building shapes, which executes the convolution directly on the points of the buildings without constructing the graphs and extracting the geometric features of the points. A new convolution operator named TriangleConv was designed to learn the feature representations of each point by aggregating the features of the point and the local triangle constructed by the point and its two adjacency points. The proposed method was evaluated and compared with related methods based on a dataset consisting of 5010 vector buildings. In terms of accuracy, macro-precision, macro-recall, and macro-F1, the results show that the proposed method has comparable performance with typical graph neural networks of GCN, GAT, and GraphSAGE, and point cloud neural networks of PointNet, PointNet++, and DGCNN in the task of recognizing and classifying building shapes in map space.
Highlights
As an indispensable part of geographic objects, buildings are widely distributed in large and medium-sized maps [1,2]
Many methods have been presented to extract the geometric features of the buildings from different perspectives for the cognition of the building shapes, such as the smallest bounding rectangle (SBR) [9] and the triangular centroid distances (TCDs) [10]
The main contribution of this paper is that we propose a new convolution operator TriangleConv, which can execute the convolution directly on the points of the buildings and build a deep point convolutional network to classify the shapes of the buildings in map space
Summary
As an indispensable part of geographic objects, buildings are widely distributed in large and medium-sized maps [1,2]. Related work has been done to extract the deep feature representations of the buildings and recognize their shapes based on graph neural networks. This paper–different from these existing works–proposes a deep point convolutional network (DPCN) to learn the deep shape feature representations of the buildings in map space and recognize their shapes. To evaluate the proposed method, these vector polygon data representing the buildings that have been used in the work of [28] are taken as the experimental dataset. The main contribution of this paper is that we propose a new convolution operator TriangleConv, which can execute the convolution directly on the points of the buildings and build a deep point convolutional network to classify the shapes of the buildings in map space.
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
