URBAN LAND COVER MAPPING USING RANDOM FOREST COMBINED WITH OPTICAL AND SAR DATA

被引:16
作者
Zhang, Hongsheng [1 ]
Zhang, Yuanzhi [1 ]
Lin, Hui [1 ]
机构
[1] Chinese Univ Hong Kong, Inst Space & Earth Informat Sci, Shatin, Hong Kong, Peoples R China
来源
2012 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS) | 2012年
关键词
Random Forest; SAR; Fusion; Classification; MARKOV RANDOM-FIELD; CLASSIFICATION; FUSION;
D O I
10.1109/IGARSS.2012.6352600
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Accurate land covers classification is challenging in urban areas due to the diversity of urban land covers. This study presents a classification strategy with combined optical and Synthetic Aperture Radar (SAR) images using Random Forest (RF). Optimization of RF is conducted, indicating the optimal number of decision trees is 10 and the optimal number of features is 4 for splitting each tree node. The overall accuracy (OA) and Kappa coefficient are used to assess the classification. Result shows that classification with combined optical and SAR images (OA: 69.08%; Kappa: 0.6288) is higher than that with single optical image (OA: 81.43%; Kappa: 0.7770). Benefits of the combined use of optical and SAR images mainly come from reducing the confusions between water and shade, and between bare soil and dark impervious surfaces.
引用
收藏
页码:6809 / 6812
页数:4
相关论文
共 20 条
[1]   NEURAL NETWORK APPROACHES VERSUS STATISTICAL-METHODS IN CLASSIFICATION OF MULTISOURCE REMOTE-SENSING DATA [J].
BENEDIKTSSON, JA ;
SWAIN, PH ;
ERSOY, OK .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 1990, 28 (04) :540-552
[2]   Random forests [J].
Breiman, L .
MACHINE LEARNING, 2001, 45 (01) :5-32
[3]   CONTRIBUTION OF AIRBORNE FULL-WAVEFORM LIDAR AND IMAGE DATA FOR URBAN SCENE CLASSIFICATION [J].
Chehata, Nesrine ;
Guo, Li ;
Mallet, Clement .
2009 16TH IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, VOLS 1-6, 2009, :1669-+
[4]   Random Forests for land cover classification [J].
Gislason, PO ;
Benediktsson, JA ;
Sveinsson, JR .
PATTERN RECOGNITION LETTERS, 2006, 27 (04) :294-300
[5]   Relevance of airborne lidar and multispectral image data for urban scene classification using Random Forests [J].
Guo, Li ;
Chehata, Nesrine ;
Mallet, Clement ;
Boukir, Samia .
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 2011, 66 (01) :56-66
[6]   Investigation of the random forest framework for classification of hyperspectral data [J].
Ham, J ;
Chen, YC ;
Crawford, MM ;
Ghosh, J .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2005, 43 (03) :492-501
[7]   TEXTURAL FEATURES FOR IMAGE CLASSIFICATION [J].
HARALICK, RM ;
SHANMUGAM, K ;
DINSTEIN, I .
IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS, 1973, SMC3 (06) :610-621
[8]   Mapping invasive plants using hyperspectral imagery and Breiman Cutler classifications (RandomForest) [J].
Lawrence, RL ;
Wood, SD ;
Sheley, RL .
REMOTE SENSING OF ENVIRONMENT, 2006, 100 (03) :356-362
[9]   Use of impervious surface in urban land-use classification [J].
Lu, Dengsheng ;
Weng, Qihao .
REMOTE SENSING OF ENVIRONMENT, 2006, 102 (1-2) :146-160
[10]   Land Cover Classification in a Complex Urban-Rural Landscape with QuickBird Imagery [J].
Lu, Dengsheng ;
Hetrick, Scott ;
Moran, Emilio .
PHOTOGRAMMETRIC ENGINEERING AND REMOTE SENSING, 2010, 76 (10) :1159-1168