Beijing University of Civil Engineering and Architecture, Beijing 100044, China.
Sensors (Basel). 2019 Dec 5;19(24):5378. doi: 10.3390/s19245378.
Tomographic synthetic aperture radar (TomoSAR) produces 3-D point clouds with unavoidable noise or false targets that seriously deteriorate the quality of 3-D images and the building reconstruction over urban areas. In this paper, a Hough transform was adopted to detect the outline of a building; however, on one hand, the obtained outline of a building with Hough transform is broken, and on the other hand, some of these broken lines belong to the same segment of a building outline, but the parameters of these lines are slightly different. These problems will lead to that segment of a building outline being represented by multiple different parameters in the Hough transform. Therefore, an unsupervised clustering method was employed for clustering these line parameters. The lines gathered in the same cluster were considered to correspond to a same segment of a building outline. In this way, different line parameters corresponding to a segment of a building outline were integrated into one and then the continuous outline of the building in cloud points was obtained. Steps of the proposed data processing method were as follows. First, the Hough transform was made use of to detect the lines on the tomography plane in TomoSAR point clouds. These detected lines lay on the outline of the building, but they were broken due to the density variation of point clouds. Second, the lines detected using the Hough transform were grouped as a date set for training the building outline. Unsupervised clustering was utilized to classify the lines in several clusters. The cluster number was automatically determined via the unsupervised clustering algorithm, which meant the number of straight segments of the building edge was obtained. The lines in each cluster were considered to belong to the same straight segment of the building outline. Then, within each cluster, which represents a part or a segment of the building edge, a repaired straight line was constructed. Third, between each two clusters or each two segments of the building outline, the joint point was estimated by extending the two segments. Therefore, the building outline was obtained as completely as possible. Finally, taking the estimated building outline as the clustering center, supervised learning algorithm was used to classify the building cloud point and the noise (or false targets), then the building cloud point was refined. Then, our refined and unrefined data were fed into the neural network for building the 3-D construction. The comparison results show the correctness and the effectiveness of our improved method.
层析合成孔径雷达（TomoSAR）生成的 3-D 点云存在不可避免的噪声或虚假目标，这会严重降低 3-D 图像的质量和城市地区的建筑物重建质量。在本文中，采用霍夫变换检测建筑物的轮廓；但是，一方面，霍夫变换得到的建筑物轮廓是破碎的，另一方面，这些破碎线中的一些属于同一建筑物轮廓的线段，但这些线的参数略有不同。这些问题将导致建筑物轮廓的线段在霍夫变换中用多个不同的参数表示。因此，采用无监督聚类方法对这些线参数进行聚类。聚类中聚集的线被认为对应于建筑物轮廓的同一线段。这样，对应于建筑物轮廓的不同线参数被集成到一个参数中，然后从云点中获得建筑物的连续轮廓。所提出的数据处理方法的步骤如下。首先，在 TomoSAR 点云中利用霍夫变换检测层析平面上的线。这些检测到的线位于建筑物的轮廓上，但由于点云的密度变化而断裂。其次，将霍夫变换检测到的线分组为一个数据集，用于训练建筑物轮廓。利用无监督聚类对这些线进行分类，将它们分为几个聚类。聚类的数量通过无监督聚类算法自动确定，这意味着获得了建筑物边缘的直线段的数量。每个聚类中的线被认为属于建筑物轮廓的同一直线段。然后，在每个聚类中，构建一个修复后的直线段，该直线段代表建筑物边缘的一部分或一段。接下来，在建筑物轮廓的两个聚类或两个线段之间，通过延伸两个线段来估计连接点。这样，就可以尽可能完整地获取建筑物轮廓。最后，以估计的建筑物轮廓作为聚类中心，利用有监督学习算法对建筑物云点和噪声（或虚假目标）进行分类，然后对建筑物云点进行细化。然后，将我们细化和未细化的数据输入神经网络中，以构建 3-D 建筑物。比较结果表明了我们改进方法的正确性和有效性。
