To explore the identification method of tree species in planted forests based on the Bayesian optimization convolutional neural network (PCA-BO-CNN) algorithm model based on principal component analysis, to improve the accuracy and robustness of remote sensing technology in tree species identification in planted forests. In this study, the Sahanba Mechanical Forest Farm was selected as the study area. Sentinel-1 remote sensing data, Sentinel-2 remote sensing data, DEM data, and the forest resource category 2 survey data were combined with the PCA-BO-CNN algorithm model, and compared with other algorithm models, to improve the accuracy of tree species identification in planted forests. The results showed that: (1) Compared with the pre-PCA algorithm, the standard deviation of 39 features of PCA1-PCA39 of multi-source data features after PCA algorithm processing and the differentiation among features were significantly improved. Therefore, PCA was beneficial to improve the identification accuracy of the dominant species of Larix gmelinii var. principis-rupprechtii (Mayr) Pilger, Betula platyphylla Sukaczev, Pinus sylvestris var. mongholica Litv., Quercus mongolica Fisch. ex Ledeb. and Picea asperata Mast. as well as non-forest land. (2) Before PCA algorithm, the overall accuracy (OA) and Kappa coefficient accuracy of the BO-random forest (BO-RF) algorithm model for the identification of dominant tree species and non-forest land were 81.87% and 0.754 5, respectively. After PCA algorithm processing, the accuracy of OA and Kappa coefficients of the PCA-BO-CNN algorithm model for the identification of dominant tree species and non-forest land was relatively improved, which were 83.10% and 0.770 3, respectively. (3) Compared with the BO-RF algorithm model before PCA algorithm processing, the overall accuracy of the PCA-BO-CNN algorithm model after PCA algorithm processing was relatively higher for the identification of the F1, OA and Kappa coefficients of the main dominant tree species and non-forest land in Saihanba Forest Farm. Specifically, compared with the BO-RF algorithm model, the OA of the PCA-BO-CNN algorithm model increased by 1.24%. In addition, the OA of the PCA-BO-CNN algorithm model before PCA algorithm was improved by 3.71%. Compared with other algorithm models, the tree species identification method based on the PCA-BO-CNN algorithm model had strong accuracy and robustness, which can help us grasp the tree species distribution of the planted forests in Saihanba Forest Farm. Further, it provides an important methodological and theoretical basis for understanding forest carbon storage, forest response to climate change, formulating carbon emission reduction policies, and promoting forest sustainable development.
MAX W, XIONGK N, ZHANGY,et al.Research progresses and prospects of carbon storage in forest ecosystems[J].Journal of Northwest Forestry University,2019,34(5):62-72.
ZHANGY, MENGN, JIANGY F.Coupling and long-term change characteristics analysis of forest carbon sequestration and forestry economic development in China[J].Journal of Beijing Forestry University,2022,44(10):129-141.
[5]
ZHANGD N, ZHAOY H, WUJ S.Assessment of carbon balance attribution and carbon storage potential in China's terrestrial ecosystem[J].Resources,Conservation and Recycling,2023,189(2):106748.
TUH T, ZHOUH B, MAG Q,et al.Characteristics of forest carbon storage in Yunnan based on the ninth forest inventory data[J].Journal of Northwest Forestry University,2023,38(3):185-193.
LIH Y, CHENY F, CHENQ,et al.Research progress of forest tree species identification based on remote sensing technology[J].Journal of Northwest Forestry University,2021,36(6):220-229.
LIUJ Z, WANGX F, WANGT.Image recognition of tree species based on multi feature fusion and CNN model[J].Journal of Beijing Forestry University,2019,41(11):76-86.
[20]
周维勋.基于深度学习特征的遥感影像检索研究[J].测绘学报,2023,52(1):167-179.
[21]
ZHOUW X.Research on remote sensing image retrieval based on deep learning features[J].Acta Geodaetica et Cartographica Sinica,2023,52(1):167-179.
LIH Y, HUANGZ P, ZHANGG L,et al.Quality assessment of CNN hyper-parameters based on dynamic weight evidential reasoning rule[J].Journal of Chinese Computer Systems,2021,42(5):1015-1021.
BARSHANE, GHODSIA, AZIMIFARZ,et al.Supervised principal component analysis: Visualization,classification and regression on subspaces and submanifolds[J].Pattern Recognition,2011,44(7):1357-1371.
WANGY B, LIY Y, HANQ,et al.Gas emission prediction of the stope in coal mine based on PCA-BO-XGBoost[J].Journal of Xi'an University of Science and Technology,2022,42(2):371-379.
FANY Y, ZHANGS S.Hyperspectral remote sensing image classification method based on deep active learning[J].Journal of Northeast Normal University (Natural Science Edition),2022,54(4):64-70.
[40]
胡越,罗东阳,花奎,关于深度学习的综述与讨论[J].智能系统学报,2019,14(1):1-19.
[41]
HUY, LUOD Y, HUAK,et al.Overview on deep learning[J].CAAI Transactions on Intelligent Systems,2019,14(1):1-19.
LIUL, ZHANGJ L, HANX L,et al.Dominant species classification based on Google Earth Engine and Sentinel time-series data[J].Forest Engineering,2023,39(1):63-72.
LIB, LIC G, LIY.Research on larch extraction in Saihanba Mechanical Forest Farm based on Sentinel-2 data[J].Forest Resource Management,2021,1(2):117-123.