Spatial distribution information extraction of macadamia forest based on GF-2 remote sensing image
-
摘要: 【目的】 基于GF-2遥感影像快速准确获取澳洲坚果林的空间分布信息，为有效利用GF-2遥感影像研究西南山区澳洲坚果林分布及为山地丘陵区其他地物类型信息的提取提供参考依据。【方法】 以云南省临沧市镇康县南伞镇为研究区，GF-2影像和数字高程模型（DEM）为数据源。通过面向对象的方法，提取影像对象的光谱特征、纹理特征、形状特征和地形特征共90维特征变量，设计8种组合方案（方案A1～方案A8），使用平均不纯度减少的方法对特征重要性进行度量，选取最佳特征组合，采用随机森林、支持向量机和决策树算法对澳洲坚果林进行提取，探讨不同特征类型和分类算法对澳洲坚果林提取精度的影响。【结果】 相比遍历分割参数法，尺度参数估算（ESP）工具和邻域差分绝对值与标准差比（RMAS）结合的方法能够更高效、客观地确定特定地物的最佳分割尺度；通过对比方案A8和方案A7可知，方案A8中加入地形特征后，整体特征维度有所降低，主要表现为纹理特征数量减少，仅保留4个纹理特征。不同类型特征对澳洲坚果林识别的贡献排序为光谱特征>地形特征>纹理特征>形状特征。在分类算法角度方面，随机森林在总体精度（OA）、用户精度（UA）、生产者精度（PA）和Kappa系数等精度指标上均优于支持向量机和决策树，方案A8融合了所有特征取得最佳的分类效果，4个指标均高于其他方案。光谱特征、纹理特征、形状特征和地形特征组合的随机森林分类方法精度最佳，OA达95.8%，澳洲坚果林的PA为87.7%,UA为94.3%。澳洲坚果林空间分布特征结果显示，澳洲坚果在15°～20°坡度范围的种植面积最大，为2.9 km2；澳洲坚果林面积主要分布在东南坡向和900～1200 m的海拔范围内。【结论】 地形+纹理+形状+地形组合方案经特征优选后结合随机森林算法，能够有效识别澳洲坚果林的分布。GF-2遥感数据与面向对象法在南方山地丘陵区澳洲坚果林制图与资源监测具有应用潜力，可用于该地区其他地物类型信息的识别。Abstract: 【Objective】 This study aimed to quickly and accurately obtain the spatial distribution information of macadamia forests based on GF-2 remote sensing image. It provided reference for the effective utilization of GF-2 remote sensing image to study the distribution of macadamia forests in the southwestern mountainous areas, as well as for the extraction of other land cover types in hilly and mountainous regions. 【Method】 The study area was located in Nansan Town, Zhenkang County, Lincang City, Yunnan Province. GF-2 image and digital elevation model（DEM） were used as data sources. An object-oriented approach was employed to extract 90 dimensional feature variables, including spectral, texture, shape and terrain features. Eight feature combination schemes（A1 to A8） were designed. The importance of the features was measured using the mean decrease in impurity（MDI） method, and the best feature combination was selected. Random forest（RF）, support vector machine（SVM）, and decision tree（DT） algorithms were used for the extraction of macadamia nut forests. The study explored the influence of different feature types and classification algorithms on the accuracy of macadamia nut forest extraction.【Result】Compared to the exhaustive segmentation parameter method, the combination of the scale parameter estimation（ESP） tool and the neighborhood difference absolute value and standard deviation ratio（RMAS） method was more efficient and objective in determining the optimal segmentation scale for specific land cover types. By comparing scheme A8 with scheme A7, it was found that adding terrain as a feature in scheme A8 reduced the overall feature dimensionality, particularly in the texture features, with only 4 texture features retained. The contribution of different feature types to the macadamia nut forest identification was ranked as follows: spectral features ＞ terrain features ＞ texture features ＞ shape features. In terms of classification algorithms, random forest outperformed support vector machine and decision tree in overall accuracy（OA）, user accuracy（UA）, producer accuracy（PA） and Kappa coefficient. Scheme A8, which integrated all features, achieved the best classification results, all higher than those of other schemes. Among the combinations of spectral, texture, shape, and terrain features, the random forest classification method achieved the best accuracy, with an OA of 95.8%, PA of 87.7%, and UA of 94.3%. The spatial distribution of macadamia nut forests showed that the largest plantation area was in the slope range of 15°-20°, covering 2.9 km2. The macadamia nut forest area was mainly distributed in the southeast-facing slopes and at altitudes of 900-1200 m. 【Conclusion】 The combination scheme of terrain, texture, shape, and terrain after feature selection along with the random forest algorithm can effectively identify the distribution of macadamia nut forests. GF-2 remote sensing data and the object-oriented method have potential applications for mapping and resource monitoring of macadamia forests in the southern mountainous hilly regions and can be used for the identification of other land cover types in the region.
-
Keywords:
- macadamia /
- GF-2 remote sensing image /
- object-oriented /
- feature optimization /
- random forest
-
-
崔宾阁,吴景,李心慧,任广波,路燕.2023.结合深度学习和植被指数的滨海湿地高分二号遥感影像信息提取[J].遥感学报,27(6):1376-1386.[Cui B G,Wu J,Li X H,Ren G B,Lu Y.2023.Combination of deep learning and vegetation index for coastal wetland mapping using GF-2remote sensing images[J].National Remote Sensing Bulle-tin,27(6):1376-1386.]doi: 10.11834/jrs.20221658. 冯林艳,谭炳香,刘清旺,周超凡,于航,张会儒,符利勇.2022.基于GF-2影像的崇礼冬奥核心区土地覆盖和树种分类[J].林业科学,58(10):10-23.[Feng L Y,Tan B X,Liu Q W,Zhou C F,Yu H,Zhang H R,Fu L Y.2022.Land cover and tree species classification of the Chongli Winter Olympic core area based on GF-2 images[J].Scientia Sil-vae Sinicae,58(10):10-23.]doi: 10.11707/j.1001-7488.20221002. 侯蒙京,殷建鹏,葛静,李元春,冯琦胜,梁天刚.2020.基于随机森林的高寒湿地地区土地覆盖遥感分类方法[J].农业机械学报,51(7):220-227.[Hou M J,Yin J P,Ge J,Li YC,Fen Q S,Liang T G.2020.Land cover remote sensing classification method of alpine wetland region based on random forest algorithm[J].Transactions of the Chinese Society for Agricultural Machinery,51(7):220-227.]doi: 10.6041/j.issn.1000-1298.2020.07.025. 黄建文,李增元,陈尔学,赵磊,莫冰萍.2021.高分六号宽幅多光谱数据人工林类型分类[J].遥感学报,25(2):539-548.[Huang J W,Li Z Y,Chen E X,Zhao L,Mo B P.2021.Classification of plantation types based on WFVmultispectral imagery of the GF-6 satellite[J].National Remote Sensing Bulletin,25(2):539-548.]doi: 10.11834/jrs.20219090. 黄邵东,徐伟恒,熊源,吴超,代飞,徐海峰,王雷光,寇卫利.2021.结合纹理和空间特征的多光谱影像面向对象茶园提取[J].光谱学与光谱分析,41(8):2565-2571.[Huang S D,Xu W H,Xiong Y,Wu C,Dai F,Xu H F,Wang L G,Kou W L.2021.Combining textures and spatial features to extract tea plantations based on object-oriented method by using multispectral image[J].Spectroscopy and Spectral Analysis,41(8):2565-2571.]doi: 10.3964/j.issn.1000-0593(2021)08-2565-07. 金梦婷,徐权,郭鹏,韩宝华,金军.2023.基于面向对象多特征学习的无人机影像农作物精细分类方法[J].遥感技术与应用,38(3):588-598.[Jin M T,Xu Q,Guo P,Han BH,Jin J.2023.Crop classification method from UAVimages based on object-oriented multi-feature learning[J].Remote Sensing Technology and Application,38(3):588-598.]doi: 10.11873/i.issn.1004-0323.2023.3.0588. 李华康,吴金龙,车雄,陈鹏鑫.2022.基于SWOT分析的镇康澳洲坚果产业发展战略研究[J].广东蚕业,56 (12):71-73.[Li H K,Wu J L,Che X,Chen P X.2022.Research on Zhenkang macadamia nut industry development strategy based on SWOT analysis[J].Guangdong Sericulture,56(12):71-73.]doi: 10.3969/j.issn.2095-1205.2022.12.20. 刘绍贵,姬忠林,张月平,李文西,高晖,杭天文,陈明,颜怡,姜义,吴兵,龚鑫鑫,祝飘,任红艳.2017.基于GF-1影像面向对象分类方法的水稻种植信息提取研究[J].中国稻米,23(6):43-46.[Liu S G,Ji Z L,Zhang Y P,Li W X,Gao H,Hang T W,Chen M,Yan Y,Jiang Y,Wu B,Gong X X,Zhu P,Ren H Y.2017.Planting information extrac-tion of rice by object-oriented classification method based on GF-1 images[J].China Rice,23(6):43-46.]doi: 10.3969/j.issn.1006-8082.2017.06.008. 刘姚,何楠,曾小姗,吴世军.2023.关于澳洲坚果研究的文献分析及发展趋势[J].中国南方果树,52(3):247-249.[Liu Y,He N,Zeng X S,Wu S J.2023.Literature analysis and future development of macadamia nut research[J].South China Fruits,52(3):247-249.]doi: 10.13938/j.issn.1007-1431.20220231. 罗红霞,戴声佩,李茂芬,李海亮,胡盈盈,郑倩,禹萱.2024.基于Gaofen-2影像和面向对象的椰子林分类研究[J].热带作物学报,45(5):1021-1030.[Luo H X,Dai S P,Li MF,Li H L,Hu Y Y,Zheng Q,Yu X.2024.Object-oriented classification of coconut palms based on Gaofen-2 imagery[J].Chinese Journal of Tropical Crops,45(5):1021-1030.]doi: 10.3969/j.issn.1000-2561.2024.05.017. 吕昱,范燕敏,武红旗,彭田田,皇甫蓓炯,贺梦婕.2020.面向对象的作物种植信息提取研究--以新疆奇台县为例[J].山东农业科学,52(6):137-143.[LüY,Fan Y M,Wu H Q,Peng T T,Huangfu B J,He M J.2020.Study on extracting crop planting information by object-oriented method taking Qitai County of Xinjiang as example[J].Shandong Agricultural Sciences,52(6):137-143.]doi: 10.14083/j.issn.1001-4942.2020.06.024. 孟浩然,李存军,郑翔宇,宫雨生,刘玉,潘瑜春.2023.综合光谱纹理和时序信息的油茶遥感提取研究[J].光谱学与光谱分析,43(5):1589-1597.[Meng H R,Li C J,Zheng XY,Gong Y S,Liu Y,Pan Y C.2023.Research on extrac-tion of Camellia oleifera by integrating spectral,texture-and time sequence remote sensing information[J].Spec-troscopy and Spectral Analysis,43(5):1589-1597.]doi: 10.3964/j.issn.1000-0593(2023)05-1589-09. 钱瑞,徐伟恒,黄邵东,王雷光,鲁宁,欧光龙.2023.应用GF-5高光谱遥感影像提取山区茶园[J].光谱学与光谱分析,43(11):3591-3598.[Qian R,Xu W H,Huang S D,Wang L G,Lu N,Ou G L.2023.Tea plantations extraction based on GF-5 hyperspectral remote sensing imagery in the mountainous area[J].Spectroscopy and Spectral Analysis,43(11):3591-3598.]doi: 10.3964/j.issn.1000-0593(2023)11-3591-08. 汤紫霞,李蒙蒙,汪小钦,邱鹏勋.2020.基于GF-2遥感影像的葡萄大棚信息提取[J].中国农业科技导报,22(11):95-105.[Tang Z X,Li M M,Wang X Q,Qiu P X.2020.Extraction of grape greenhouses from GF-2 remote sensing images[J].Journal of Agricultural Science and Tech-nology,22(11):95-105.]doi: 10.13304/j.nykjdb.2019.0759. 吴涵敏,李金星,劳慧洁.2023.百色市老山林场澳洲坚果优质高产栽培管理技术[J].南方农业,17(16):49-51.[Wu H M,Li J X,Lao H J.2023.High quality and high yield cultivation and management techniques for macadamia nuts in Laoshan forest farm of Baise City[J].South China Agriculture,17(16):49-51.]doi: 10.19415/j.cnki.1673-890x.2023.16.016. 夏清,李建华,代硕,张涵,邢学敏.2023.顾及潮汐影响的中国红树林高分二号遥感制图[J].遥感学报,27(6):1320-1333.[Xia Q,Li J H,Dai S,Zhang H,Xing X M.2023.Mapping high-resolution mangrove forests in China using GF-2 imagery under the tide[J].National Remote Sensing Bulletin,27(6):1320-1333.]doi: 10.11834/jrs.20221848. 徐凡迪,白海东,樊绍光,万晓丽,杨建荣,李智华,苏建荣,张燕平,吴疆翀,赵云晋.2022.澳洲坚果繁育系统研究进展[J].世界林业研究,35(5):37-41.[Xu F D,Bai H D,Fan S G,Wan X L,Yang J R,Li Z H,Su J R,Zhang Y P,Wu J C,Zhao Y J.2022.Research advances in macadamia breeding system[J].World Forestry Research,35(5):37-41.]doi: 10.13348/j.cnki.sjlyyj.2022.0043.y. 徐朋,徐伟诚,罗阳帆,赵祚喜.2019.基于无人机可见光遥感影像的耕地精准分类方法研究[J].中国农业科技导报,21(6):79-86.[Xu P,Xu W C,Luo Y F,Zhao Z X.2019.Precise classification of cultivated land based on visible remote sensinc image of UAV[J].Journal of Agricultural Science and Technology,21(6):79-86.]doi: 10.13304/j.nykjdb.2019.0103. 杨丽萍,岳海,何双凌,陶亮,贺熙勇.2021.云南澳洲坚果园土壤和叶片养分评价[J].热带作物学报,42(8):2269-2274.[Yang L P,Yue H,He S L,Tao L,He X Y.2021.Evaluation of soil and leaf nutrients in macadamia orchards in Yunnan China[J].Chinese Journal of Tropical Crops,42(8):2269-2274.]doi: 10.3969/j.issn.1000-2561.2021.08.020. 张睿,马建文.2009.支持向量机在遥感数据分类中的应用新进展[J].地球科学进展,24(5):555-562.[Zhang R,Ma JW.2009.State of the art on remotely sensed data classifica-tion based on support vector machines[J].Advances in Earth Science,24(5):555-562.]doi: 10.3321/j.issn:1001-8166.2009.05.012. 张悦琦,任鸿瑞.2023.融合特征优选与随机森林算法的GF-6影像东北一季稻遥感提取[J].遥感学报,27(9):2153-2164.[Zhang Y Q,Ren H R.2023.Remote sensing extrac-tion of paddy rice in Northeast China from GF-6 images by combining feature optimization and random forest[J].National Remote Sensing Bulletin,27(9):2153-2164.]doi: 10.11834/jrs.20221338. 赵国兵,郑江华,王蕾,高健,罗磊,尼格拉·吐尔逊,韩万强,关靖云.2024.基于RF分类调优和SNIC聚类的新疆红枣种植区遥感提取[J].干旱区地理,47(6):1004-1014.[Zhao G B,Zheng J H,Wang L,Gao J,Luo L,Nigela T,Han W Q,Guan J Y.2024.Remote sensing extraction of jujube planting area in Xinjiang based on RF classification optimization and SNlC clustering[J].Arid Land Geogra-phy,47(6):1004-1014.]doi: 10.12118/j.issn.1000-6060.2023.382. 朱红春,蔡丽杰,刘海英,江涛.2015.高分辨率影像分类的最佳分割尺度计算[J].测绘科学,40(3):71-75.[Zhu H C,Cai L J,Liu H Y,Jiang T.2015.Optimal segmentation scale calculation for high-resolution remote sensing image[J].Science of Surveying and Mapping,40(3):71-75.]doi: 10.16251/j.cnki.1009-2307.2015.03.015. Breiman L.2001.Random forests[J].Machine Learning,45(1):5-32.doi: 10.1023/a:1010933404324.
Haralick R M.1979.Statistical and structural approaches to tex-ture[J].Proceedings of the IEEE,67(5):786-804.doi: 10.1109/PROC.1979.11328.
Ke Y H,Quackenbush L J,Im J H.2010.Synergistic use of Quick Bird multispectral imagery and LIDAR data for object-based forest species classification[J].Remote Sensing of Environment,114:1141-1154.doi: 10.1016/j.rse.2010.01.002.
Luo H X,Li M F,Dai S P,Li H L,Li Y P,Hu Y Y,Zheng Q,Yu X,Fang J H.2022.Combinations of feature selection and machine learning algorithms for object-oriented betel palms and mango plantations classification based on Gaofen-2 imagery[J].Remote Sensing,14(7):1757.doi: 10.3390/rs14071757.
Ma L,Fu T Y,Blaschke T,Li M C,Tiede D,Zhou Z J,Ma XX,Chen D L.2017.Evaluation of feature selection methods for object-based land cover mapping of unmanned aerial vehicle imagery using random forest and support vector Machine classifiers[J].ISPRS International Journal of Geo-Information,6(2):51.doi: 110.3390/ijgi6020051.
Shang M,Wang S X,Zhou Y,Du C,Liu W L.2018.Objectbased image analysis of suburban landscapes using landsat-8 imagery[J].International Journal of Digital Earth,12(6):720-736.doi: 10.1080/17538947.
Wang M C,Li M J,Wang F Y,Ji X.2022.Exploring the opti-mal feature combination of tree species classification by fusing multi-feature and multi-temporal sentinel-2 data in Changbai Mountain[J].Forests,13(7):1058.doi: 10.3390/f13071058.
Wu Q,Zhong R F,Zhao W J,Song K,Du L M.2019.Landcover classification using GF-2 images and airborne lidar data based on random forest[J].International Journal of Remote Sensing,40(5-6):2410-2426.doi: 10.1080/01431161.2018.1483090.
计量
- 文章访问数: 9
- HTML全文浏览量: 9
- PDF下载量: 3
下载: