Scientia Silvae Sinicae ›› 2026, Vol. 62 ›› Issue (5): 80-95.doi: 10.11707/j.1001-7488.LYKX20250332
• Research papers • Previous Articles Next Articles
Jiahao Wang1,2,3(
),Yifan Xie1,2,3,Jinghang Ge1,2,3,Jiaxin Wang1,2,3,Xiaoli Zhang1,2,3,*(
),Xin Tian4
Received:2025-05-26
Revised:2025-10-28
Online:2026-05-10
Published:2026-10-17
Contact:
Xiaoli Zhang
E-mail:wangjiahao7@bjfu.edu.cn;zhang-xl@263.net
CLC Number:
Jiahao Wang,Yifan Xie,Jinghang Ge,Jiaxin Wang,Xiaoli Zhang,Xin Tian. Small Sample Classification of Tree Species Using Combined Satellite Multispectral Imagery and Airborne LiDAR[J]. Scientia Silvae Sinicae, 2026, 62(5): 80-95.
Table 1
Sample dataset information"
| 类别 Types | 训练像元数 Training pixels | 测试像 元数 Test pixels | 合计 Total | |
| 主要树种 Main tree species | 栎类Quercus | 1 098 | 2 602 | 3 700 |
| 桉树Eucalyptus | 960 | 2 240 | 3 200 | |
| 思茅松Pinus kesiya var. langbianensis | 988 | 2 212 | 3 200 | |
| 其他树种 Other tree species | 706 | 1 594 | 2 300 | |
| 农作物 Crops | 耕地Farmland | 1 319 | 3 081 | 4 400 |
| 茶园Tea plantation | 1 267 | 3 173 | 4 440 | |
| 咖啡园Coffee plantation | 804 | 1 896 | 2 700 | |
| 非植被 Non-vegetation | 水体Water | 876 | 1 924 | 2 800 |
| 建设用地Buildings | 857 | 1 943 | 2 800 | |
| 采伐迹地Cutting-site | 587 | 1 413 | 2 000 | |
Table 2
Classification accuracy of different training set ratio on GF-6"
| 训练样本占比 Training set ratio | 20 | 40 | 60 | 80 | 100 | |||||||||||||||
| 支持集参数K-shot | 1 | 5 | 10 | 1 | 5 | 10 | 1 | 5 | 10 | 1 | 5 | 10 | 1 | 5 | 10 | |||||
| 总体精度Overall accuracy (%) | 79.51 | 80.23 | 80.56 | 80.1 | 81.63 | 81.65 | 84.7 | 84.64 | 84.40 | 78.73 | 84.54 | 83.87 | 84.04 | 85.42 | 84.15 | |||||
| 栎类 Quercus | 63.85 | 67.03 | 65.91 | 68.69 | 67.38 | 69.06 | 70.42 | 68.91 | 70.44 | 68.11 | 69.54 | 71.20 | 71.42 | 73.60 | 73.51 | |||||
| 桉树Eucalyptus | 85.33 | 81.01 | 85.75 | 85.34 | 82.1 | 84.11 | 87.17 | 88.95 | 85.43 | 92.06 | 89.38 | 92.24 | 88.51 | 88.79 | 88.45 | |||||
| 思茅松Pinus kesiya var. langbianensis | 81.58 | 81.12 | 80.84 | 86.01 | 84.24 | 88.75 | 89.79 | 86.50 | 84.29 | 80.93 | 84.55 | 82.67 | 90.91 | 88.87 | 90.34 | |||||
| 其他树种Other tree species | 48.44 | 49.59 | 49.49 | 51.3 | 55.85 | 54.08 | 61.17 | 55.91 | 58.49 | 54.63 | 57.95 | 55.5 | 56.97 | 58.03 | 55.27 | |||||
| 耕地Farmland | 84.81 | 86.18 | 86.37 | 83.14 | 87.22 | 84.72 | 95.48 | 92.97 | 93.18 | 92.88 | 92.43 | 91.29 | 92.78 | 93.23 | 93.54 | |||||
| 茶园Tea plantation | 82.78 | 80.55 | 82.60 | 88.66 | 83.76 | 86.67 | 76.18 | 81.07 | 81.99 | 79.72 | 86.08 | 86.23 | 84.54 | 88.56 | 86.23 | |||||
| 咖啡园Coffee plantation | 64.58 | 68.08 | 68.39 | 58.42 | 66.43 | 64.43 | 83.28 | 82.23 | 80.24 | 51.46 | 74.52 | 72.38 | 67.87 | 71.60 | 69.69 | |||||
| 水体Water | 99.95 | 99.95 | 99.90 | 100.00 | 100.00 | 100.00 | 100.00 | 99.74 | 100.00 | 99.79 | 100 | 99.95 | 99.84 | 100 | 99.95 | |||||
| 建设用地Buildings | 95.16 | 97.38 | 96.88 | 97.88 | 97.65 | 96.13 | 93.76 | 95.86 | 94.77 | 99.58 | 98.67 | 97.84 | 98.87 | 97.83 | 95.79 | |||||
| 采伐迹地Cutting-site | 91.52 | 89.66 | 88.96 | 95.00 | 91.09 | 92.52 | 96.28 | 95.06 | 91.50 | 90.55 | 92.82 | 93.35 | 95.30 | 94.73 | 90.97 |
Table 3
Ablation experiments of the improved prototype network on Sentinel-2 and GF-6 data"
| 数据 Data | 注意力机制 CBAM | 监督学习 SimCLR | 训练时间 Training time/s | 平均精度 Average accuracy (AA) (%) | 总体精度 Overall accuracy (OA) (%) | |
| 空间模块 Spatial | 通道模块 Channel | |||||
| Sentinel-2 | ⅹ | ⅹ | ⅹ | 166 | 92.10 | 92.11 |
| √ | ⅹ | ⅹ | 268 | 92.18 | 92.20 | |
| ⅹ | √ | ⅹ | 325 | 92.15 | 92.19 | |
| √ | √ | ⅹ | 212 | 92.31 | 92.32 | |
| √ | √ | √ | 281 | 92.50 | 92.41 | |
| GF-6 | ⅹ | ⅹ | ⅹ | 228 | 85.52 | 85.42 |
| √ | ⅹ | ⅹ | 280 | 85.58 | 85.51 | |
| ⅹ | √ | ⅹ | 270 | 85.53 | 85.40 | |
| √ | √ | ⅹ | 262 | 86.11 | 86.65 | |
| √ | √ | √ | 438 | 87.23 | 87.05 | |
Table 4
Classification accuracy of the improved prototype network for various ground object categories"
| 编号Code | 类别名称Name | 精确率Precision (%) | 召回率Recall (%) | F1分数 F1-score (%) | |||||
| Sentinel-2 | GF-6 | Sentinel-2 | GF-6 | Sentinel-2 | GF-6 | ||||
| 1 | 栎类 Quercus | 94.39 | 81.70 | 84.74 | 68.64 | 89.31 | 74.60 | ||
| 2 | 桉树Eucalyptus | 93.49 | 90.58 | 96.83 | 89.24 | 95.13 | 89.90 | ||
| 3 | 思茅松Pinus kesiya var. langbianensis | 90.41 | 92.13 | 98.46 | 91.55 | 94.27 | 91.84 | ||
| 4 | 其他树种Other tree species | 87.86 | 59.74 | 85.38 | 76.98 | 86.61 | 67.27 | ||
| 5 | 耕地Farmland | 96.87 | 94.05 | 94.42 | 90.75 | 95.63 | 92.37 | ||
| 6 | 茶园Tea plantation | 92.42 | 85.56 | 84.90 | 84.81 | 88.50 | 85.19 | ||
| 7 | 咖啡园Coffee plantation | 75.87 | 77.35 | 93.35 | 87.18 | 83.71 | 81.97 | ||
| 8 | 水体Water | 99.69 | 99.90 | 99.95 | 99.74 | 99.82 | 99.82 | ||
| 9 | 建设用地Buildings | 99.79 | 97.22 | 91.82 | 93.52 | 95.73 | 95.33 | ||
| 10 | 采伐迹地Cutting-site | 94.00 | 94.09 | 99.79 | 92.43 | 96.81 | 93.25 | ||
Table 5
Classification results of combining Sentinel-2 band combinations, GF-6 resolution combinations, and CHM"
| 数据 Data | 组合方式 Combination method | 训练时间 Training times(%) | 平均精度 Average sccuracy (AA) (%) | 总体精度 Overall accuracy (OA) (%) |
Sentinel-2 | 原始10波段 Origin 10 bands | 166 | 92.10 | 92.11 |
| 原始10波段+CHM Origin 10 bands+CHM | 228 | 95.95(+3.85) | 95.96(+3.85) | |
| 降维4波段 PCA4 bands | 108 | 84.56 | 83.99 | |
| 降维4波段+CHM PCA4 bands+CHM | 147 | 87.62(+3.06) | 87.43(+3.44) | |
| 原始2 m Origin 2 m | 228 | 85.52 | 85.42 | |
| GF-6 | 原始2 m+CHM Origin 2 m+CHM | 248 | 88.04(+2.52) | 88.15(+2.73) |
| 重采样10 m Resample 10 m | 97 | 90.42 | 90.14 | |
| 重采样10 m+CHM Resample 10 m+CHM | 101 | 93.36(+2.94) | 93.07(+2.93) |
Table 6
Classification accuracy of various ground object categories under the original prototype network combined with CHM"
| 编号Code | 类别名称Name | 精确率Precision (%) | 召回率Recall (%) | F1分数F1-score (%) | |||||
| Sentinel-2 | GF-6 | Sentinel-2 | GF-6 | Sentinel-2 | GF-6 | ||||
| 1 | 栎类 Quercus | 93.53 | 75.63 | 96.66 | 78.25 | 95.07 | 76.92 | ||
| 2 | 桉树Eucalyptus | 95.98 | 92.06 | 99.20 | 91.16 | 97.56 | 91.61 | ||
| 3 | 思茅松Pinus kesiya var. langbianensis | 98.44 | 94.00 | 97.02 | 89.87 | 97.72 | 91.89 | ||
| 4 | 其他树种Other tree species | 94.96 | 64.47 | 89.84 | 67.50 | 92.33 | 65.95 | ||
| 5 | 耕地Farmland | 97.24 | 93.83 | 98.44 | 94.84 | 97.84 | 94.33 | ||
| 6 | 茶园Tea plantation | 99.46 | 93.50 | 87.61 | 79.33 | 93.16 | 85.83 | ||
| 7 | 咖啡园Coffee plantation | 84.26 | 75.80 | 98.26 | 91.19 | 90.72 | 82.79 | ||
| 8 | 水体Water | 99.84 | 99.90 | 99.95 | 99.84 | 99.90 | 99.92 | ||
| 9 | 建设用地Buildings | 99.95 | 96.65 | 95.78 | 94.40 | 97.82 | 95.77 | ||
| 10 | 采伐迹地Cutting-site | 95.79 | 94.48 | 99.86 | 98.16 | 97.78 | 96.29 | ||
Table 7
Classification accuracy of different algorithms on Sentinel-2 and GF-6 data combined with CHM"
| 编号Code | 类别名称Name | 随机森林RF | 二维卷积2D-CNN | 本研究方法Methods used in this study | |||||
| Sentinel-2 | GF-6 | Sentinel-2 | GF-6 | Sentinel-2 | GF-6 | ||||
| 1 | 栎类 Quercus | 85.21 | 68.29 | 95.67 | 71.77 | 93.30 | 83.03 | ||
| 2 | 桉树Eucalyptus | 90.71 | 86.47 | 93.22 | 88.15 | 97.98 | 92.91 | ||
| 3 | 思茅松Pinus kesiya var. langbianensis | 94.64 | 90.71 | 95.91 | 86.83 | 98.01 | 91.20 | ||
| 4 | 其他树种Other tree species | 86.20 | 75.67 | 93.84 | 80.20 | 95.26 | 63.17 | ||
| 5 | 耕地Farmland | 91.55 | 81.39 | 98.87 | 88.22 | 98.79 | 95.41 | ||
| 6 | 茶园Tea plantation | 87.21 | 77.26 | 96.77 | 88.14 | 94.46 | 90.76 | ||
| 7 | 咖啡园Coffee plantation | 95.78 | 81.13 | 95.58 | 86.38 | 95.61 | 83.61 | ||
| 8 | 水体Water | 99.95 | 99.74 | 99.79 | 99.69 | 99.95 | 99.64 | ||
| 9 | 建设用地Buildings | 97.92 | 95.82 | 99.22 | 97.05 | 99.79 | 96.69 | ||
| 10 | 采伐迹地Cutting-site | 98.81 | 95.96 | 96.65 | 94.16 | 97.72 | 94.11 | ||
| 平均精度 Average accuracy(%) | 92.80 | 85.24 | 96.55 | 88.06 | 97.09 | 89.05 | |||
| 总体精度 Overall accuracy(%) | 91.96 | 83.60 | 96.62 | 87.34 | 96.92 | 89.21 | |||
Table 8
Classification results of various ground object categories on the improved prototype network combined with CHM"
| 编号Code | 类别名称Name | 精确率Precision (%) | 召回率Recall (%) | F1分数 F1-score (%) | |||||
| Sentinel-2 | GF-6 | Sentinel-2 | GF-6 | Sentinel-2 | GF-6 | ||||
| 1 | 栎类Quercus | 93.53 | 75.63 | 96.66 | 78.25 | 95.07 | 76.92 | ||
| 2 | 桉树Eucalyptus | 95.98 | 92.06 | 99.20 | 91.16 | 97.56 | 91.61 | ||
| 3 | 思茅松Pinus kesiya var. langbianensis | 98.44 | 94.00 | 97.02 | 89.87 | 97.72 | 91.89 | ||
| 4 | 其他树种Other tree species | 94.96 | 64.47 | 89.84 | 67.50 | 92.33 | 65.95 | ||
| 5 | 耕地Farmland | 97.24 | 93.83 | 98.44 | 94.84 | 97.84 | 94.33 | ||
| 6 | 茶园Tea plantation | 99.46 | 93.50 | 87.61 | 79.33 | 93.16 | 85.83 | ||
| 7 | 咖啡园Coffee plantation | 84.26 | 75.80 | 98.26 | 91.19 | 90.72 | 82.79 | ||
| 8 | 水体Water | 99.84 | 99.90 | 99.95 | 99.84 | 99.90 | 99.92 | ||
| 9 | 建设用地Buildings | 99.95 | 96.65 | 95.78 | 94.40 | 97.82 | 95.77 | ||
| 10 | 采伐迹地Cutting-site | 95.79 | 94.48 | 99.86 | 98.16 | 97.78 | 96.29 | ||
Table 9
Statistics of overlapping pixels of main tree species"
| 类别名称 Name | Sentinel-2重叠 像元占比 Percentage of overlapping pixels (Sentinel-2)(%) | GF-6重叠 像元占比 Percentage of overlapping pixels (Sentinel-2)(%) |
| 栎类 Quercus | 70.67 | 73.87 |
| 桉树Eucalyptus | 62.75 | 63.56 |
| 思茅松Pinus kesiya var. langbianensis | 62.45 | 61.69 |
| 其他树种Other tree species | 49.39 | 46.70 |
| 耕地Farmland | 72.18 | 70.07 |
| 茶园Tea plantation | 77.94 | 70.75 |
| 咖啡园Coffee plantation | 64.83 | 63.50 |
| 李树涛, 李聪妤, 康旭东. 多源遥感图像融合发展现状与未来展望. 遥感学报, 2021, 25 (1): 148- 166. | |
| Li S T, Li C Y, Kang X D. Development status and future prospects of multi-source remote sensing image fusion. Journal of Remote Sensing, 2021, 25 (1): 148- 166. | |
|
李 煜, 李崇贵, 刘思涵, 等. 应用哨兵2A多时相遥感影像对树种的识别. 东北林业大学学报, 2021, 49 (3): 44- 47,51.
doi: 10.3969/j.issn.1000-5382.2021.03.008 |
|
|
Li Y, Li C G, Liu S H, et al. Tree species recognition with sentinel-2A multitemporal remote sensing image. Journal of Northeast Forestry University, 2021, 49 (3): 44- 47,51.
doi: 10.3969/j.issn.1000-5382.2021.03.008 |
|
| 李增元, 刘清旺, 庞 勇. 激光雷达森林参数反演研究进展. 遥感学报, 2016, 20 (5): 1138- 1150. | |
| Li Z Y, Liu Q W, Pang Y. Review on forest parameters inversion using LiDAR. Journal of Remote Sensing, 2016, 20 (5): 1138- 1150. | |
| 刘大伟, 韩 玲, 韩晓勇. 基于深度学习的高分辨率遥感影像分类研究. 光学学报, 2016, 36 (4): 298- 306. | |
| Liu D W, Han L, Han X Y. High spatial resolution remote sensing image classification based on deep learning. Acta Optica Sinica, 2016, 36 (4): 298- 306. | |
| 刘 戈, 姜小光, 唐伯惠. 特征优选与卷积神经网络在农作物精细分类中的应用研究. 地球信息科学学报, 2021, 23 (6): 1071- 1081. | |
| Liu G, Jiang S G, Tang B H. Application of feature optimization and convolutional neural network in crop classification. Journal of Geo-Information Science, 2021, 23 (6): 1071- 1081. | |
|
刘鲁霞, 庞 勇, 李增元. 基于地基激光雷达的亚热带森林单木胸径与树高提取. 林业科学, 2016, 52 (2): 26- 37.
doi: 10.11707/j.1001-7488.20160204 |
|
|
Liu L X, Pang Y, Li Z Y. Individual tree DBH and height estimation using terrestrial laser scanning (TLS) in a subtropical forest. Scientia Silvae Sinica, 2016, 52 (2): 26- 37.
doi: 10.11707/j.1001-7488.20160204 |
|
|
刘世荣, 王 晖, 李海奎, 等. 碳中和目标下中国森林碳储量、碳汇变化预估与潜力提升途径. 林业科学, 2024, 60 (4): 157- 172.
doi: 10.11707/j.1001-7488.LYKX20230206 |
|
|
Liu S R, Wang H, Li H K, et al. Projections of China’s forest carbon storage and sequestration and ways of their potential capacity enhancement. Scientia Silvae Sinica, 2024, 60 (4): 157- 172.
doi: 10.11707/j.1001-7488.LYKX20230206 |
|
|
刘 颖, 雷研博, 范九伦, 等. 基于小样本学习的图像分类技术综述. 自动化学报, 2021, 47 (2): 297- 315.
doi: 10.16383/j.aas.c190720 |
|
|
Liu Y, Lei Y B, Fan J L, et al. Survey on image classification technology based on small sample learning. Acta Automatica Sinica, 2021, 47 (2): 297- 315.
doi: 10.16383/j.aas.c190720 |
|
| 马鸿伟, 刘 海, 姚顺彬, 等. 基于林业遥感的树种分类应用分析与展望. 林业资源管理, 2020 (3): 118- 121. | |
| Ma H W, Liu H, Yao S B, et al. Analysis and prospect of tree species classification based on forestry remote sensing. Forest Resources Management, 2020 (3): 118- 121. | |
|
王美华, 吴振鑫, 周祖光. 基于注意力改进CBAM的农作物病虫害细粒度识别研究. 农业机械学报, 2021, 52 (4): 239- 247.
doi: 10.6041/j.issn.1000-1298.2021.04.025 |
|
|
Wang M H, Wu Z X, Zhou Z G. Fine-grained identification research of crop pests and diseases based on improved CBAM via attention. Transactions of the Chinese Society for Agricultural Machinery, 2021, 52 (4): 239- 247.
doi: 10.6041/j.issn.1000-1298.2021.04.025 |
|
|
王智超, 马天天, 邵亚奎, 等. 面向未来的中国智慧林业: 观测仪器体系的演进与发展趋势. 林业科学, 2024, 60 (4): 1- 15.
doi: 10.11707/j.1001-7488.LYKX20220903 |
|
|
Wang Z C, Ma T T, Shao Y K, et al. Future oriented smart forestry in China: evolution and development trends of observation instrument systems. Scientia Silvae Sinicae, 2024, 60 (4): 1- 15.
doi: 10.11707/j.1001-7488.LYKX20220903 |
|
| 肖庆琳, 张加龙, 曹 军, 等. 耦合多特征多时相的普洱市优势树种分类研究. 森林工程, 2024, 40 (2): 117- 126. | |
| Xiao Q L, Zhang J L, Cao J, et al. 2024. Research on the classification of dominant tree species in Pu’er City by coupling multiple characteristics and multiple phases. Forest Engineering, 2024, 40 (2): 117- 126. | |
|
张良培, 沈焕锋. 遥感数据融合的进展与前瞻. 遥感学报, 2016, 20 (5): 1050- 1061.
doi: 10.11834/jrs.20166243 |
|
|
Zhang, L P, and Shen H F. Progress and future of remote sensing data fusion. Journal of Remote Sensing, 2016, 20 (5): 1050- 1061.
doi: 10.11834/jrs.20166243 |
|
| 朱钟正, 苏 伟. 基于局部空间统计分析的SPOT 5影像分类. 遥感学报, 2011, 15 (5): 957- 972. | |
| Zhu Z Z, Su W. The analysis of the classification of SPOT 5 image based on local spatial statistics. Journal of Remote Sensing, 2011, 15 (5): 957- 972. | |
|
左溪冰, 刘 智, 金 飞, 等. 面向高光谱影像小样本分类的全局−局部特征自适应融合方法. 地球信息科学学报, 2023, 25 (8): 1699- 1716.
doi: 10.12082/dqxxkx.2023.230058 |
|
|
Zuo X B, Liu Z, Jin F, et al. Global-local feature adaptive fusion method for small sample classification of hyperspectral images. Journal of Geo-Information Science, 2023, 25 (8): 1699- 1716.
doi: 10.12082/dqxxkx.2023.230058 |
|
| Alajaji D, Alhichri H S, Ammour N, et al. 2020. Few-shot learning for remote sensing scene classification. 2020 Mediterranean and Middle-East Geoscience and Remote Sensing Symposium (M2GARSS), March 9−11, 2020, Tunis, Tunisia, IEEE, 81−84. | |
|
Chakraborty D, Ciceu A, Ballian D, et al. Assisted tree migration can preserve the European forest carbon sink under climate change. Nature Climate Change, 2024, 14 (8): 845- 852.
doi: 10.1038/s41558-024-02080-5 |
|
|
Chen L, Wei Y X, Yao Z Q, et al. Data augmentation in prototypical networks for forest tree species classification using airborne hyperspectral images. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60, 1- 16.
doi: 10.1109/tgrs.2022.3168054 |
|
|
Chen L, Wu J, Xie Y F, et al. Discriminative feature constraints via supervised contrastive learning for few-shot forest tree species classification using airborne hyperspectral images. Remote Sensing of Environment, 2023, 295, 113710.
doi: 10.1016/j.rse.2023.113710 |
|
|
Chen L, Tian X M, Chai G Q, et al. A new CBAM-P -net model for few-shot forest species classification using airborne hyperspectral images. Remote Sensing, 2021, 13 (7): 1269.
doi: 10.3390/rs13071269 |
|
|
de Conto T, Armston J, Dubayah R. Characterizing the structural complexity of the earth’s forests with spaceborne LiDAR. Nature Communications, 2024, 15, 8116.
doi: 10.1038/s41467-024-52468-2 |
|
|
Deur M, Gašparović M, Balenović I. Tree species classification in mixed deciduous forests using very high spatial resolution satellite imagery and machine learning methods. Remote Sensing, 2020, 12 (23): 3926.
doi: 10.3390/rs12233926 |
|
| Jiang D W, Li W B, Cao M, et al. 2020. Speech SIMCLR: combining contrastive and reconstruction objective for self-supervised speech representation learning. arXiv: 2010.13991. https://arxiv.org/abs/2010.13991. | |
| Hamzaoui M, Chapel L, Pham M T, et al. 2022. A hierarchical prototypical network for few-shot remote sensing scene classification. Pattern Recognition and Artificial Intelligence. Cham: Springer International Publishing, 208−220. | |
| Zhang H Y, Cisse M, Dauphin Y N, et al. 2017. Mixup: beyond empirical risk minimization. arXiv: 1710.09412. https://arxiv.org/abs/1710.09412. | |
| Inoue H. 2018. Data augmentation by pairing samples for images classification. arXiv: 1801.02929. https://arxiv.org/abs/1801.02929. | |
|
Li Y S, Zhu Z H, Yu J G, et al. Learning deep cross-modal embedding networks for zero-shot remote sensing image scene classification. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59 (12): 10590- 10603.
doi: 10.1109/TGRS.2020.3047447 |
|
|
Lei L, Chai G, Yao Z Q, et al. A novel self-similarity cluster grouping approach for individual tree crown segmentation using multi-features from UAV-based LiDAR and multi-angle photogrammetry data. Remote Sensing of Environment, 2025, 318, 114588.
doi: 10.1016/j.rse.2024.114588 |
|
|
Qiu C P, Zhang X Y, Tong X C, et al. Few-shot remote sensing image scene classification: recent advances, new baselines, and future trends. ISPRS Journal of Photogrammetry and Remote Sensing, 2024, 209, 368- 382.
doi: 10.1016/j.isprsjprs.2024.02.005 |
|
|
Ramezan C A, Warner T A, Maxwell A E, et al. Effects of training set size on supervised machine-learning land-cover classification of large-area high-resolution remotely sensed data. Remote Sensing, 2021, 13 (3): 368.
doi: 10.3390/rs13030368 |
|
| Woo S, Park J, Lee J Y, et al. 2018. CBAM: convolutional block attention module. Computer Vision – ECCV 2018. Cham: Springer International Publishing, 3−19. | |
| Snell J, Swersky K, Zemel R. 2017. Prototypical networks for few-shot learning. arXiv: 1703.05175. https://arxiv.org/abs/1703.05175. | |
| DeVries T, Taylor G W. 2017. Improved regularization of convolutional neural networks with cutout. arXiv: 1708.04552. https://arxiv.org/abs/1708.04552. | |
| Yun S, Han D, Chun S, et al. 2019. CutMix: regularization strategy to train strong classifiers with localizable features. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), October 27-November 2, 2019, Seoul, Korea, IEEE, 6022−6031. | |
|
Wu J, Chen L, Wang J Q, et al. A novel framework combining band selection algorithm and improved 3D prototypical network for tree species classification using airborne hyperspectral images. Computers and Electronics in Agriculture, 2024, 219, 108813.
doi: 10.1016/j.compag.2024.108813 |
| [1] | Guangsheng Chen,Linzhi Wen,Wenjun Zhang,Chao Li,Ming Yu,Weipeng Jing. A Cross-Domain Generalization Classification Model for Tree Species in Complex Scenes Based on Feature Fusion [J]. Scientia Silvae Sinicae, 2025, 61(4): 33-45. |
| [2] | Bingbing Yang,Jie Xu. Tree Species Recognition Based on Improved ConvNeXt Network [J]. Scientia Silvae Sinicae, 2025, 61(2): 31-39. |
| [3] | Zhongyang Hu,Liang Shan,Xiangyu Chen,Kunyong Yu,Jian Liu. Individual Tree Segmentation of UAV-LiDAR Based on the Combination of CHM and DSM [J]. Scientia Silvae Sinicae, 2024, 60(8): 14-24. |
| [4] | Yingwu Mao,Ying Guo,Wangfei Zhang,Yong Su,Yuan Guan. Tree Species Classification by Combining LiDAR, Hyperspectral Data and 3D-CNN Method [J]. Scientia Silvae Sinicae, 2023, 59(3): 73-83. |
| [5] | Jiajie Su,Zheyu Zhang,Jiajun Xu,Bin Li,Jun Lü,Qing Yao. Forest Pest Identification Method Based on a Deep Bilinear Transformation Attention Mechanism Network [J]. Scientia Silvae Sinicae, 2023, 59(2): 121-128. |
| [6] | Linyan Feng,Bingxiang Tan,Qingwang Liu,Chaofan Zhou,Hang Yu,Huiru Zhang,Liyong Fu. Land Cover and Tree Species Classification of the Chongli Winter Olympic Core Area Based on GF-2 Images [J]. Scientia Silvae Sinicae, 2022, 58(10): 10-23. |
| [7] | Dongbo Xie,Yakai Lei,Yuchao Zhang,Qingwang Liu,Liyong Fu,Qiao Chen. Estimation of Canopy Cover in the Core Area of Winter Olympic Games Based on Airborne LiDAR Data [J]. Scientia Silvae Sinicae, 2022, 58(10): 24-34. |
| [8] | Lin Zhao,Xiaoli Zhang,Yanshuang Wu,Bin Zhang. Subtropical Forest Tree Species Classification Based on 3D-CNN for Airborne Hyperspectral Data [J]. Scientia Silvae Sinicae, 2020, 56(11): 97-107. |
| [9] | Liu Jinli, Chen Zhao, Gao Jinping, Gao Xianlian, Sun Zhongqiu. Research on the Method of Determining the Optimal Segmentation Scale for Tree Species Classification of High-Resolution Image [J]. Scientia Silvae Sinicae, 2019, 55(11): 95-104. |
| [10] | Duan Zhugeng, Xiao Huashun, Yuan Weixiang. Comparison of Interpolation Methods of Forest Canopy Height Model Using Discrete Point Cloud Data [J]. Scientia Silvae Sinicae, 2016, 52(9): 86-94. |
| Viewed | ||||||
|
Full text |
|
|||||
|
Abstract |
|
|||||