A Multi-Temporal and Multi-Feature Larch Plantation Extraction Study Based on GF-1 Images

doi:10.13466/j.cnki.lyzygl.2022.04.014

Abstract

Abstract:

Larch plantation has become an important tree species in north China,and the afforestation area is increasing year by year.Therefore,the extraction of larch plantation is of great significance for rational utilization of forest resources in China.Based on Mengjiagang Forest Farm within the territory of Huanan County in Heilongjiang Province as the research district,combined with phenological characteristics of larch,the study selected a typical period of GF-1 PMS images as the data source,used forest resources subcompartment data and field survey data as the sample data,extracted image spectral characteristics,texture characteristics,vegetation index and topographic features,and used Random Forest (RF)algorithm to extract the spatial distribution of larch plantation from multi-date and multi-feature angles,so as to obtain the best classification time phase and feature combination of larch plantation.The experimental results showed that the optimal window size was 9×9 by using gray co-occurrence matrix (GLCM)to classify texture features under different windows.The importance of all features was evaluated based on the Gini coefficient,and the highest overall accuracy was selected as the preferred subset.When 84% of all features (spectral,texture,exponential and topographic features were 11,5,9 and 2 respectively)were used for classification,the overall accuracy reached the highest 82.67% (Kappa coefficient was 0.76).The contribution rate of the vegetation index was the highest among all the characteristics.Compared with spectral features,spectral features+ vegetation index,spectral features + texture features and spectral features + terrain factors classification,constructing multi-feature optimization RF classification model can effectively reduce the dimension and improve the classification accuracy of larch plantationforest.

Key words: larch plantation forest, texture features, random forest, feature preference

CLC Number:

S771.8

WANG Xiaoyang, JIANG Youyi, LI Xiao, HU Yaxuan, ZHANG Jiazheng, LIU Bowei. A Multi-Temporal and Multi-Feature Larch Plantation Extraction Study Based on GF-1 Images[J]. FOREST RESOURCES WANAGEMENT, 2022, (4): 109-118.

Figures/Tables 8

Fig.1

Fig.2

Tab.1

Fig.3

Fig.4

Fig.5

Fig.6

Tab.2

References 33

[1]	郭瑞霞. 基于多源数据的落叶松人工林识别研究[D]. 西安: 西安科技大学, 2019.
[2]	Gao Tian, Zhu Jiaojun, Zheng Xiao, et al. Mapping spatial distribution of larch plantations from multi-seasonal Landsat-8 OLI imagery and multi-scale textures using random forests[J]. Remote Sens, 2015, 7(2):1702-1720. doi: 10.3390/rs70201702
[3]	任冲. 中高分辨率遥感影像森林类型精细分类与森林资源变化监测技术研究[D]. 北京: 中国林业科学研究院, 2016.
[4]	周小成, 郑磊, 黄洪宇. 基于多特征优选的无人机可见光遥感林分类型分类[J]. 林业科学, 2021, 57(6):24-36.
[5]	Xu Kaijian, Tian Qingjiu, Zhang Zhaoying, et al. Tree species(genera)identification with GF-1 time-series in a forested landscape,Northeast China[J]. Remote Sens, 2020, 12(10):1554. doi: 10.3390/rs12101554
[6]	Medina Machín A, Marcello J, Hernández-Cordero A I, et al. Vegetation species mapping in a coastal-dune ecosystem using high resolution satellite imagery[J]. GIScience & Remote Sensing, 2019, 56(2):210-232.
[7]	Abdel-Hamid A, Dubovyk O, Abou El-Magd I, et al. Mapping mangroves extents on the Red Sea coastline in Egypt using polarimetric SAR and high resolution optical remote sensing data[J]. Sustainability, 2018, 10(3):646. doi: 10.3390/su10030646
[8]	Wang Minye, Fei Xianyun, Zhang Yuanzhi, et al. Assessing texture features to classify coastal wetland vegetation from high spatial resolution imagery using completed local binary patterns(CLBP)[J]. Remote Sens, 2018, 10(5):778. doi: 10.3390/rs10050778
[9]	陈玲, 郝文乾, 高德亮. 光学影像纹理信息在林业领域的最新应用进展[J]. 北京林业大学学报, 2015, 37(3):1-12.
[10]	白莹, 胡淑萍. 基于CART决策树的自然保护区植被类型分布研究[J]. 北京林业大学学报, 2020, 42(6):113-122.
[11]	Deur M, Gasparovic M, Balenovic I. Tree species classification in mixed deciduous forests using very high spatial resolution satellite imagery and machine learning methods[J]. Remote Sens, 2020, 12(23):3926. doi: 10.3390/rs12233926
[12]	Lu Dengsheng, Batistella M, De Miranda E E, et al. A comparative study of Landsat TM and SPOT HRG images for vegetation classification in the Brazilian Amazon[J]. Photogrammetric engineering and remote sensing, 2008, 74(3):311-321. doi: 10.14358/pers.74.3.311 pmid: 19789716
[13]	Chen Zhou, Fei Xianyun, Gao Xiangwei, et al. The influence of CLBP window size on urban vegetation type classification using high spatial resolution satellite images[J]. Remote Sens, 2020, 12(20):3393. doi: 10.3390/rs12203393
[14]	张嵩, 马保东, 陈玉腾, 等. 融合遥感影像光谱和纹理特征的矿区林地信息变化监测[J]. 地理与地理信息科学, 2017, 33(6):44-49.
[15]	王猛, 张新长, 王家耀, 等. 结合随机森林面向对象的森林资源分类[J]. 测绘学报, 2020, 49(2):235-244.
[16]	刘怀鹏, 安慧君, 王冰, 等. 基于递归纹理特征消除的WorldView-2树种分类[J]. 北京林业大学学报, 2015, 37(8):53-59.
[17]	夏盈, 厉恩华, 王学雷, 等. 基于特征优选的随机森林算法在湿地信息提取中的应用——以湖北洪湖湿地自然保护区为例[J]. 华中师范大学学报:自然科学版, 2021, 55(4):639-648.
[18]	李哲, 张沁雨, 彭道黎. 基于高分二号遥感影像的树种分类方法[J]. 遥感技术与应用, 2019, 34(5):970-982.
[19]	杨蕙宇. 顾及特征类型和特征选择的面向对象森林分类方法对比研究[D]. 西安: 陕西师范大学, 2019.
[20]	Haralick R M, Shanmugam K, Dinstein I. Textural features for image classification[J]. IEEE Transactions on Systems,Man and Cybernetics, 1973, 3(6):610-621.
[21]	徐辉. 基于Landsat-8 OLI和GF-2遥感影像的森林资源识别研究[D]. 南昌: 江西农业大学, 2018.
[22]	刘思涵, 尚夏明, 马婷. 基于随机森林特征选择的森林类型分类[J]. 北京测绘, 2019, 33(12):1518-1522.
[23]	Breiman L. Random forests[J]. Machine Learning, 2001, 45(1):5-32. doi: 10.1023/A:1010933404324
[24]	窦世卿, 宋莹莹, 徐勇, 等. 基于随机森林的高分影像分类及土地利用变化检测[J]. 无线电工程, 2021, 51(9):901-908.
[25]	林海晏, 岳彩荣, 吴晓晖, 等. 基于EnMAP-Box的遥感图像分类研究[J]. 西南林业大学学报, 2014, 34(2):67-71.
[26]	林雪, 彭道黎, 黄国胜, 等. 结合多尺度纹理特征的遥感影像面向对象分类[J]. 测绘工程, 2016, 25(7):22-27.
[27]	杨盼盼. 基于高分辨率遥感影像纹理特征的面向对象植被分类方法研究[D]. 昆明: 云南师范大学, 2017.
[28]	Ma Minfei, Liu Jianhong, Liu Mingxing, et al. Tree species classification based on Sentinel-2 imagery and random forest classifier in the eastern regions of the Qilian Mountains[J]. Forests, 2021, 12(12):1736. doi: 10.3390/f12121736
[29]	吕杰, 郝宁燕, 李崇贵, 等. 利用随机森林和纹理特征的森林类型识别[J]. 遥感信息, 2017, 32(6):109-114.
[30]	孙玉琳, 潘洁. 基于Landsat 8影像的南京市紫金山风景林区树种分类研究[J]. 国土与自然资源研究, 2022(3):64-68.
[31]	何云, 黄翀, 李贺, 等. 基于Sentinel-2A影像特征优选的随机森林土地覆盖分类[J]. 资源科学, 2019, 41(5):992-1001. doi: 10.18402/resci.2019.05.15
[32]	王李娟, 孔钰如, 杨小冬, 等. 基于特征优选随机森林算法的农耕区土地利用分类[J]. 农业工程学报, 2020, 36(4):244-250.
[33]	张红华, 赵威成, 刘强凯. 特征优选随机森林的土地利用分类[J]. 黑龙江科技大学学报, 2020, 30(5):490-494.

分类方案	特征	特征个数/个
1	光谱特征	12
2	光谱特征+植被指数	21
3	光谱特征+纹理特征	20
4	光谱特征+地形因子	15
5	光谱特征+植被指数+纹理特征+地形因子	32
6	优选特征子集	27

分类方案	总体精度/%	Kappa系数	F精度/%	生产者精度/%	用户精度/%
1	79.57	0.7227	85.26	88.37	82.36
2	79.92	0.7289	85.99	87.49	84.55
3	79.52	0.7209	85.06	86.24	83.88
4	78.89	0.7121	82.84	84.42	81.32
5	82.02	0.7560	86.18	88.72	83.79
6	82.67	0.7648	87.13	89.70	84.70