arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.09300cs.CVeess.IV
收藏

数据量对基于深度学习的超高空间分辨率航空影像建筑屋顶提取性能的影响

The impact of data volume on performance of deep learning based building rooftop extraction using very high spatial resolution aerial images

University of Waterloo(滑铁卢大学)
Hongjie He, Ke Yang, Yuwei Cai, Zijian Jiang, Qiutong Yu, Kun Zhao, Junbo Wang, Sarah Narges Fatholahi, Yan Liu, Hasti Andon Petrosians, Bingxu Hu, Liyuan Qing,… 展开作者

Hongjie He, Ke Yang, Yuwei Cai, Zijian Jiang, Qiutong Yu, Kun Zhao, Junbo Wang, Sarah Narges Fatholahi, Yan Liu, Hasti Andon Petrosians, Bingxu Hu, Liyuan Qing, Zhehan Zhang, Hongzhang Xu, Siyu Li, Kyle Gao, Linlin Xu, Jonathan Li

摘要:

建筑物屋顶数据在一些城市应用和自然灾害管理中非常重要。与传统测绘相比,通过使用高空间分辨率的航空影像,基于深度学习的建筑物屋顶提取方法高效准确。虽然在基于深度学习的任务中更倾向于使用更多的训练数据,但数据量对构建提取模型的影响尚未得到充分研究。因此,本文探讨了数据量对使用基于深度学习的方法从超高空间分辨率(VHSR)图像中提取建筑物屋顶的性能的影响。为此,我们手动标记了0.12米空间分辨率的航空图像,并使用流行的深度学习架构对不同大小的数据集进行了训练,以执行分割任务,包括全卷积网络(FCN)-8s,U-Net和DeepLabv 3+。实验表明,随着训练数据的增加,算法收敛速度加快,精度提高,而更好的算法能够更好地缓解训练数据不足的问题。

英文摘要:

Building rooftop data are of importance in several urban applications and in natural disaster management. In contrast to traditional surveying and mapping, by using high spatial resolution aerial images, deep learning-based building rooftops extraction methods are efficient and accurate. Although more training data is preferred in deep learning-based tasks, the effect of data volume on building extraction models is underexplored. Therefore, the paper explores the impact of data volume on the performance of building rooftop extraction from very-high-spatial-resolution (VHSR) images using deep learning-based methods. To do so, we manually labelled 0.12m spatial resolution aerial images and perform a comparative analysis of models trained on datasets of different sizes using popular deep learning architectures for segmentation tasks, including Fully Convolutional Networks (FCN)-8s, U-Net and DeepLabv3+. The experiments showed that with more training data, algorithms converged faster and achieved higher accuracy, while better algorithms were able to better mitigate the lack of training data.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑