Abstract

This paper investigates the deep neural networks for rapid and accurate detection of building rooftops in aerial orthoimages. The networks were trained using the manually labeled rooftop vector data digitized on aerial orthoimagery covering the Kitchener-Waterloo area. The performance of the three deep learning methods, U-Net, Fully Convolutional Network (FCN), and Deeplabv3+ were compared by training, validation, and testing sets in the dataset. Our results demonstrated that DeepLabv3+ achieved 63.8% in Intersection over Union (IoU), 77.8% in mean IoU (mIoU), 74% in precision, and 78% in F₁-score. After improving the performance with focal loss, training loss was greatly cut down and the convergence rate experienced a significant growth. Meanwhile, rooftop detection also achieved higher performance, as Deeplabv3+ reached 93.6% in average pixel accuracy, with 65.4% in IoU, 79.0% in mIoU, 77.6% in precision, and 79.1% in F₁-score. Lastly, in order to evaluate the effects of data volume, by changing data volume from 100% to 75% and 50% in ablation study, it shows that when data volume decreased, the performance of extraction also got worse, with IoU, mIoU, precision, and F₁-score also mostly decreased.

摘要

本文研究了用于在航空正射影像中快速准确检测建筑物屋顶的深度神经网络。这些网络使用手动标记的屋顶矢量数据进行训练，这些数据在覆盖基奇纳-滑铁卢地区的航空正射影像上数字化。通过数据集中的训练、验证和测试集比较了三种深度学习方法 U-Net、全卷积网络 (FCN) 和 Deeplabv3+ 的性能。我们的结果表明，DeepLabv3+ 在联合交集 (IoU) 上实现了 63.8%，在平均 IoU (mIoU) 上实现了 77.8%，在精度上实现了 74%，在 F _{1 上实现了}78%-分数。在使用focal loss提高性能后，训练损失大大减少，收敛速度显着增长。同时，屋顶检测也取得了更高的性能，Deeplabv3+的平均像素精度达到了93.6%，IoU为65.4%，mIoU为79.0%，精度为77.6%，F ₁ -score为79.1% 。最后，为了评估数据量的影响，将消融研究中的数据量从 100% 改为 75% 和 50%，表明当数据量减少时，提取性能也变差，IoU、mIoU、精度，F ₁ -score 也大多下降。