TY - JOUR
T1 - Simultaneous extraction of roads and buildings in remote sensing imagery with convolutional neural networks
AU - Alshehhi, Rasha
AU - Marpu, Prashanth Reddy
AU - Woon, Wei Lee
AU - Mura, Mauro Dalla
N1 - Publisher Copyright:
© 2017 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS)
PY - 2017/8
Y1 - 2017/8
N2 - Extraction of man-made objects (e.g., roads and buildings) from remotely sensed imagery plays an important role in many urban applications (e.g., urban land use and land cover assessment, updating geographical databases, change detection, etc). This task is normally difficult due to complex data in the form of heterogeneous appearance with large intra-class and lower inter-class variations. In this work, we propose a single patch-based Convolutional Neural Network (CNN) architecture for extraction of roads and buildings from high-resolution remote sensing data. Low-level features of roads and buildings (e.g., asymmetry and compactness) of adjacent regions are integrated with Convolutional Neural Network (CNN) features during the post-processing stage to improve the performance. Experiments are conducted on two challenging datasets of high-resolution images to demonstrate the performance of the proposed network architecture and the results are compared with other patch-based network architectures. The results demonstrate the validity and superior performance of the proposed network architecture for extracting roads and buildings in urban areas.
AB - Extraction of man-made objects (e.g., roads and buildings) from remotely sensed imagery plays an important role in many urban applications (e.g., urban land use and land cover assessment, updating geographical databases, change detection, etc). This task is normally difficult due to complex data in the form of heterogeneous appearance with large intra-class and lower inter-class variations. In this work, we propose a single patch-based Convolutional Neural Network (CNN) architecture for extraction of roads and buildings from high-resolution remote sensing data. Low-level features of roads and buildings (e.g., asymmetry and compactness) of adjacent regions are integrated with Convolutional Neural Network (CNN) features during the post-processing stage to improve the performance. Experiments are conducted on two challenging datasets of high-resolution images to demonstrate the performance of the proposed network architecture and the results are compared with other patch-based network architectures. The results demonstrate the validity and superior performance of the proposed network architecture for extracting roads and buildings in urban areas.
KW - Adjacent regions
KW - Convolutional neural network
KW - Extraction
KW - Low-level features
UR - http://www.scopus.com/inward/record.url?scp=85020312124&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85020312124&partnerID=8YFLogxK
U2 - 10.1016/j.isprsjprs.2017.05.002
DO - 10.1016/j.isprsjprs.2017.05.002
M3 - Article
AN - SCOPUS:85020312124
SN - 0924-2716
VL - 130
SP - 139
EP - 149
JO - ISPRS Journal of Photogrammetry and Remote Sensing
JF - ISPRS Journal of Photogrammetry and Remote Sensing
ER -