[TPAMI-2023] Learning to Extract Building Footprints From Off-Nadir Aerial Images

论文阅读 [TPAMI-2023] Learning to Extract Building Footprints From Off-Nadir Aerial Images

论文搜索(studyai.com)

搜索论文: Learning to Extract Building Footprints From Off-Nadir Aerial Images

搜索论文: http://www.studyai.com/search/whole-site/?q=Learning+to+Extract+Building+Footprints+From+Off-Nadir+Aerial+Images&fr=csdn

关键字(Keywords)

Buildings; Head; Training; Annotations; Task analysis; Imaging; Image segmentation; Building footprint extraction; building detection; learning offset vector; off-nadir aerial image

机器视觉

航空图像

摘要(Abstract)

Extracting building footprints from aerial images is essential for precise urban mapping with photogrammetric computer vision technologies.

从航空图像中提取建筑足迹对于利用摄影测量计算机视觉技术精确绘制城市地图至关重要。

Existing approaches mainly assume that the roof and footprint of a building are well overlapped, which may not hold in off-nadir aerial images as there is often a big offset between them.

现有的方法主要假设建筑物的屋顶和占地面积很好地重叠，这可能无法保持在最低点的空中图像中，因为它们之间通常存在较大的偏移。

In this paper, we propose an offset vector learning scheme, which turns the building footprint extraction problem in off-nadir images into an instance-level joint prediction problem of the building roof and its corresponding “roof to footprint” offset vector.

在本文中，我们提出了一种偏移向量学习方案，该方案将非最低点图像中的建筑足迹提取问题转化为建筑屋顶及其对应的“屋顶到足迹”偏移向量的实例级联合预测问题。

Thus the footprint can be estimated by translating the predicted roof mask according to the predicted offset vector.

因此，可以通过根据预测的偏移矢量平移预测的屋顶掩模来估计足迹。

We further propose a simple but effective feature-level offset augmentation module, which can significantly refine the offset vector prediction by introducing little extra cost.

我们进一步提出了一个简单但有效的特征级偏移增强模块，它可以通过引入很少的额外成本来显著细化偏移向量预测。

Moreover, a new dataset, Buildings in Off-Nadir Aerial Images (BONAI), is created and released in this paper.

此外，本文还创建并发布了一个新的数据集，即Off-Nadir航空图像中的建筑物（BONAI）。

It contains 268,958 building instances across 3,300 aerial images with fully annotated instance-level roof, footprint, and corresponding offset vector for each building.

它包含3300幅空中图像中的268958个建筑实例，具有完全注释的实例级屋顶、迹线和每个建筑的相应偏移矢量。

Experiments on the BONAI dataset demonstrate that our method achieves the state-of-the-art, outperforming other competitors by 3.37 to 7.39 points in F1-score.

在BONAI数据集上的实验表明，我们的方法达到了最先进的水平，在F1得分上超过其他竞争对手3.37至7.39分。

The codes, datasets, and trained models are available at https://github.com/jwwangchn/BONAI.git…

代码、数据集和训练模型可在https://github.com/jwwangchn/BONAI.git.。

作者(Authors)

[‘Jinwang Wang’, ‘Lingxuan Meng’, ‘Weijia Li’, ‘Wen Yang’, ‘Lei Yu’, ‘Gui-Song Xia’]