TY - GEN
T1 - Collaborative Classification for Woodland Data Using Similar Multi-concentrated Network
AU - Zhu, Yixuan
AU - Zhang, Mengmeng
AU - Li, Wei
AU - Tao, Ran
AU - Ran, Qiong
N1 - Publisher Copyright:
© 2020, Springer Nature Switzerland AG.
PY - 2020
Y1 - 2020
N2 - With the increasing of the forest area and complexity of tree species, collaborative classification using multi-source remote sensing data has been drawn increasing attention. Fusion of hyperspectral and LiDAR data can improve to acquire a comprehensive information which is conductive to the forest land classification. In this work, a similar multi-concentrate network focusing on the fine classification of tree species, denoted as SMCN, is proposed for woodland data. More specific, a preprocessing stage named pixel screening for data intensity critical control is firstly designed. Then, a similar multi-concentrate network is developed to capture spectral and spatial features from hyperspectral and LiDAR data and make specific connections, respectively. Experimental results validated on Belgian data have favorably demonstrated that the proposed SMCN outperforms other state-of-the-art methods.
AB - With the increasing of the forest area and complexity of tree species, collaborative classification using multi-source remote sensing data has been drawn increasing attention. Fusion of hyperspectral and LiDAR data can improve to acquire a comprehensive information which is conductive to the forest land classification. In this work, a similar multi-concentrate network focusing on the fine classification of tree species, denoted as SMCN, is proposed for woodland data. More specific, a preprocessing stage named pixel screening for data intensity critical control is firstly designed. Then, a similar multi-concentrate network is developed to capture spectral and spatial features from hyperspectral and LiDAR data and make specific connections, respectively. Experimental results validated on Belgian data have favorably demonstrated that the proposed SMCN outperforms other state-of-the-art methods.
KW - Collaborative classification
KW - Convolutional neural network
KW - Multi-source remote sensing data
KW - Woodland classification
UR - http://www.scopus.com/inward/record.url?scp=85093978825&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-60639-8_8
DO - 10.1007/978-3-030-60639-8_8
M3 - Conference contribution
AN - SCOPUS:85093978825
SN - 9783030606381
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 95
EP - 101
BT - Pattern Recognition and Computer Vision - 3rd Chinese Conference, PRCV 2020, Proceedings
A2 - Peng, Yuxin
A2 - Zha, Hongbin
A2 - Liu, Qingshan
A2 - Lu, Huchuan
A2 - Sun, Zhenan
A2 - Liu, Chenglin
A2 - Chen, Xilin
A2 - Yang, Jian
PB - Springer Science and Business Media Deutschland GmbH
T2 - 3rd Chinese Conference on Pattern Recognition and Computer Vision, PRCV 2020
Y2 - 16 October 2020 through 18 October 2020
ER -
