CNN-Based Individual Tree Species Classification Using High-Resolution Satellite Imagery and Airborne LiDAR Data

被引:34
作者
Li, Hui [1 ,2 ,3 ]
Hu, Baoxin [4 ]
Li, Qian [4 ]
Jing, Linhai [1 ,2 ,3 ]
机构
[1] Int Res Ctr Big Data Sustainable Dev Goals, Beijing 100094, Peoples R China
[2] Chinese Acad Sci, Aerosp Informat Res Inst, Beijing 100094, Peoples R China
[3] Hainan Key Lab Earth Observat, Sanya 572029, Peoples R China
[4] York Univ, Dept Earth & Space Sci & Engn, 4700 Keele St, Toronto, ON M3J 1P3, Canada
基金
中国国家自然科学基金;
关键词
deep learning; high-resolution satellite images; LiDAR; tree species classification; WORLDVIEW-2; IMAGERY; FOREST; UTILITY; CROWNS; NDVI; LEAF;
D O I
10.3390/f12121697
中图分类号
S7 [林业];
学科分类号
0829 ; 0907 ;
摘要
Deep learning (DL) has shown promising performances in various remote sensing applications as a powerful tool. To explore the great potential of DL in improving the accuracy of individual tree species (ITS) classification, four convolutional neural network models (ResNet-18, ResNet-34, ResNet-50, and DenseNet-40) were employed to classify four tree species using the combined high-resolution satellite imagery and airborne LiDAR data. A total of 1503 samples of four tree species, including maple, pine, locust, and spruce, were used in the experiments. When both WorldView-2 and airborne LiDAR data were used, the overall accuracies (OA) obtained by ResNet-18, ResNet-34, ResNet-50, and DenseNet-40 were 90.9%, 89.1%, 89.1%, and 86.9%, respectively. The OA of ResNet-18 was increased by 4.0% and 1.8% compared with random forest (86.7%) and support vector machine (89.1%), respectively. The experimental results demonstrated that the size of input images impacted on the classification accuracy of ResNet-18. It is suggested that the input size of ResNet models can be determined according to the maximum size of all tree crown sample images. The use of LiDAR intensity image was helpful in improving the accuracies of ITS classification and atmospheric correction is unnecessary when both pansharpened WorldView-2 images and airborne LiDAR data were used.
引用
收藏
页数:22
相关论文
共 50 条
[1]   Urban tree species mapping using hyperspectral and lidar data fusion [J].
Alonzo, Michael ;
Bookhagen, Bodo ;
Roberts, Dar A. .
REMOTE SENSING OF ENVIRONMENT, 2014, 148 :70-83
[2]  
[Anonymous], 2014, Comput. Sci.
[3]   Object-based fusion for urban tree species classification from hyperspectral, panchromatic and nDSM data [J].
Aval, Josselin ;
Fabre, Sophie ;
Zenou, Emmanuel ;
Sheeren, David ;
Fauvel, Mathieu ;
Briottet, Xavier .
INTERNATIONAL JOURNAL OF REMOTE SENSING, 2019, 40 (14) :5339-5365
[4]  
BAUER ME, 1994, PHOTOGRAMM ENG REM S, V60, P287
[5]   Effects of atmospheric correction and pansharpening on LULC classification accuracy using WorldView-2 imagery [J].
Lin, Chinsu ;
Wu, Chao-Cheng ;
Tsogt, Khongor ;
Ouyang, Yen-Chieh ;
Chang, Chein-I .
Information Processing in Agriculture, 2015, 2 (01) :25-36
[6]   Assessing the utility WorldView-2 imagery for tree species mapping in South African subtropical humid forest and the conservation implications: Dukuduku forest patch as case study [J].
Cho, Moses Azong ;
Malahlela, Oupa ;
Ramoelo, Abel .
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION, 2015, 38 :349-357
[7]   Hyperspectral discrimination of tropical rain forest tree species at leaf to crown scales [J].
Clark, ML ;
Roberts, DA ;
Clark, DB .
REMOTE SENSING OF ENVIRONMENT, 2005, 96 (3-4) :375-398
[8]  
Deering DW, 1978, THESIS TEXAS A M U C, P338
[9]   Comparison of Tree Species Classifications at the Individual Tree Level by Combining ALS Data and RGB Images Using Different Algorithms [J].
Deng, Songqiu ;
Katoh, Masato ;
Yu, Xiaowei ;
Hyyppa, Juha ;
Gao, Tian .
REMOTE SENSING, 2016, 8 (12)
[10]   Phenological differences in Tasseled Cap indices improve deciduous forest classification [J].
Dymond, CC ;
Mladenoff, DJ ;
Radeloff, VC .
REMOTE SENSING OF ENVIRONMENT, 2002, 80 (03) :460-472