Instance segmentation based building extraction in a dense urban area using multispectral aerial imagery data

被引:1
作者
Fatty, Abdoulie [1 ]
Li, An-Jui [1 ]
Yao, Chih-Yuan [2 ]
机构
[1] Natl Taiwan Univ Sci & Technol, Dept Civil & Construct Engn, Taipei 106335, Taiwan
[2] Natl Taiwan Univ Sci & Technol, Dept Comp Sci & Informat Engn, Taipei 106335, Taiwan
关键词
Instance segmentation; Deep learning; Transfer learning; Building extraction; Aerial imagery;
D O I
10.1007/s11042-023-15905-w
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Fast and automatic extraction of buildings from aerial imagery is essential for various applications. However, it remains a challenge, mainly because it requires the accurate recovery of buildings from high-resolution aerial imagery data. The task is particularly challenging due to the distinct nature of building shapes. The extraction of buildings from aerial imagery is often manually conducted by humans, which makes the process very time-consuming and costly. Furthermore, it is especially challenging to automatically extract the footprints of closely spaced buildings in dense urban areas through traditional segmentation techniques. This study proposes an instance segmentation-based building extraction technique for effective building detection and segmentation. The proposed method combines transfer learning with Mask Regional Convolutional Neural Network (Mask R-CNN) integrated with PointRend to locate and generate high-quality segmentation masks for building instances. In addition, several data augmentation strategies are also implemented to enlarge the training dataset. Results indicate that the proposed instance segmentation-based method for building extraction is able to execute distinctly efficient and effective automatic segmentation of distinctive and complex-shaped buildings in a dense urban area.
引用
收藏
页码:61913 / 61928
页数:16
相关论文
共 22 条
  • [1] Bischke B, 2019, IEEE IMAGE PROC, P1480, DOI [10.1109/ICIP.2019.8803050, 10.1109/icip.2019.8803050]
  • [2] A small attentional YOLO model for landslide detection from satellite remote sensing images
    Cheng, Libo
    Li, Jia
    Duan, Ping
    Wang, Mingguo
    [J]. LANDSLIDES, 2021, 18 (08) : 2751 - 2765
  • [3] Deng J, 2009, PROC CVPR IEEE, P248, DOI 10.1109/CVPRW.2009.5206848
  • [4] Building extraction from high-resolution satellite images in urban areas: recent methods and strategies against significant challenges
    Ghanea, Mohsen
    Moallem, Payman
    Momeni, Mehdi
    [J]. INTERNATIONAL JOURNAL OF REMOTE SENSING, 2016, 37 (21) : 5234 - 5248
  • [5] Fast R-CNN
    Girshick, Ross
    [J]. 2015 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2015, : 1440 - 1448
  • [6] Rich feature hierarchies for accurate object detection and semantic segmentation
    Girshick, Ross
    Donahue, Jeff
    Darrell, Trevor
    Malik, Jitendra
    [J]. 2014 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2014, : 580 - 587
  • [7] He KM, 2020, IEEE T PATTERN ANAL, V42, P386, DOI [10.1109/ICCV.2017.322, 10.1109/TPAMI.2018.2844175]
  • [8] Fastai: A Layered API for Deep Learning
    Howard, Jeremy
    Gugger, Sylvain
    [J]. INFORMATION, 2020, 11 (02)
  • [9] Kirillov A, 2020, PROC CVPR IEEE, P9796, DOI 10.1109/CVPR42600.2020.00982
  • [10] Li JC, 2017, INT CONF ACOUST SPEE, P126, DOI 10.1109/ICASSP.2017.7952131