TY - JOUR
T1 - Semantic Façade Segmentation from Airborne Oblique Images
AU - Lin, Yaping
AU - Nex, F.C.
AU - Yang, Michael Ying
PY - 2019/6/1
Y1 - 2019/6/1
N2 - In this paper, oblique airborne images with very high resolution are used to address the problem from aerial views in urban areas. Traditional classification method (i.e., random forests) is compared with state-of-the-art fully convolutional networks (FCNs). Random forests use hand-craft image features including red, green, blue (RGB), scale-invariant feature transform (SIFT), and Texton, and point cloud features consisting of normal vector and planarity extracted from different scales. In contrast, the inputs of FCNs are the RGB bands and the third components of normal vectors. In both cases, three-dimensional (3D) features are projected back into the image space to support the facade interpretation. Fully connected conditional random field (CRF) is finally taken as a post-processing of the FCN to refine the segmentation results. Several tests have been performed and the achieved results show that the models embedding the 3D component outperform the solution using only images. FCNs significantly outperformed random forests, especially for the balcony delineation.
AB - In this paper, oblique airborne images with very high resolution are used to address the problem from aerial views in urban areas. Traditional classification method (i.e., random forests) is compared with state-of-the-art fully convolutional networks (FCNs). Random forests use hand-craft image features including red, green, blue (RGB), scale-invariant feature transform (SIFT), and Texton, and point cloud features consisting of normal vector and planarity extracted from different scales. In contrast, the inputs of FCNs are the RGB bands and the third components of normal vectors. In both cases, three-dimensional (3D) features are projected back into the image space to support the facade interpretation. Fully connected conditional random field (CRF) is finally taken as a post-processing of the FCN to refine the segmentation results. Several tests have been performed and the achieved results show that the models embedding the 3D component outperform the solution using only images. FCNs significantly outperformed random forests, especially for the balcony delineation.
KW - ITC-ISI-JOURNAL-ARTICLE
UR - https://ezproxy2.utwente.nl/login?url=https://library.itc.utwente.nl/login/2019/isi/nex_sem.pdf
UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.14358/PERS.85.6.425
U2 - 10.14358/PERS.85.6.425
DO - 10.14358/PERS.85.6.425
M3 - Article
VL - 85
SP - 425
EP - 433
JO - Photogrammetric Engineering & Remote Sensing
JF - Photogrammetric Engineering & Remote Sensing
SN - 0099-1112
IS - 6
ER -