WANG Erli, LI Cunjun, ZHOU Jingping, PENG Dailiang, HU Haitang, DONG Xi. Classification of Beijing Afforestation Species Based on Multi-temporal Images[J]. Journal of Beijing University of Technology, 2017, 43(5): 710-718. DOI: 10.11936/bjutxb2016100039
Citation:
WANG Erli, LI Cunjun, ZHOU Jingping, PENG Dailiang, HU Haitang, DONG Xi. Classification of Beijing Afforestation Species Based on Multi-temporal Images[J]. Journal of Beijing University of Technology, 2017, 43(5): 710-718. DOI: 10.11936/bjutxb2016100039
In order to solve the problem that traditional remote sensing classification method is difficult to distinguish tree species of plain afforestation, high spatial images of four different periods were selected to present the distribution of forest resource with explicit clarity. The optimal segmentation parameters were obtained by combining the calculation of change rate in variance and visual interpretation and by using a tool named ESP. Feature selection was used to reduce a large number of features to simplify the process. The region was classified by the object-based classification by using mutli-temporal images and single image respectively. Pixel-based classifications were applied to compare with the accuracy of object-based classification method. The results show that the accuracy of object-based classification in mutli-temporal images is 64%, which is better than the single image results with the accuracy of 51%. SVM and MLC reaches even lower accuracy of 49% and 43% respectively. The precision of object-based KNN classifications is better than that of pixel-based classification, indicating that the object-based classification adding mutli-temporal images has superiority in identifying those afforestation tree species in ecological landscape of forest with a complex distribution.
Table
3.
Feature selected with multi-temproal images
最终筛选的特征
Mean_627_NIR
Mean_423_NIR
Mean_317_NIR
GLCM_Mean_423_G
Length/Width
GLCM_Cont_627_NIR
GLCM_Homo_518_NIR
GLDV_Ent_518_R
GLDV_Ent_423_G
GLCM_Cor_423_NIR
NDVI_627
Stddev_317_NIR
GLCM_Ang_423_NIR
GLCM_Mean_627_NIR
GLDV_Mean_317_NIR
Std_dev_627_NIR
注:1)表中数字代表影像日期,R、G、NIR分别代表红、绿、近红外波段;2)表中纹理特征部分采用缩写,分别为同质度(homogeneity,Homo)、对比度(contrast,Cont)、角二阶矩(angular second moment,Ang)、,熵(entropy,Ent)、非相似度(dissimilarity,Dis)、均值(mean)、标准差(stddev)、相关(correlation,Cor).
Table
3
Feature selected with multi-temproal images
最终筛选的特征
Mean_627_NIR
Mean_423_NIR
Mean_317_NIR
GLCM_Mean_423_G
Length/Width
GLCM_Cont_627_NIR
GLCM_Homo_518_NIR
GLDV_Ent_518_R
GLDV_Ent_423_G
GLCM_Cor_423_NIR
NDVI_627
Stddev_317_NIR
GLCM_Ang_423_NIR
GLCM_Mean_627_NIR
GLDV_Mean_317_NIR
Std_dev_627_NIR
注:1)表中数字代表影像日期,R、G、NIR分别代表红、绿、近红外波段;2)表中纹理特征部分采用缩写,分别为同质度(homogeneity,Homo)、对比度(contrast,Cont)、角二阶矩(angular second moment,Ang)、,熵(entropy,Ent)、非相似度(dissimilarity,Dis)、均值(mean)、标准差(stddev)、相关(correlation,Cor).
LEHMANN EA,CACCETTAP,LOWELLK,et al.SAR and optical remote sensing: assessment of complementarity and interoperability in the context of a large-scale operational forest monitoring system[J].Remote Sensing of Environment,2015,156:335-348.
[2]
PIPPURII,SUVANTOA,MALTAMOM,et al.Classification of forest land attributes using multi-source remotely sensed data[J].International Journal of Applied Earth Observation and Geoinformation,2016,44:11-22.
[3]
AGARWALS,VAILSHERY LS,JAGANMOHANM,et al.Mapping urban tree species using very high resolution satellite imagery: comparing pixel-based and object-based approaches[J].ISPRS International Journal of Geo-Information,2013,2(1):220-236.
WANGJ,ZHAO TZ,ZENGY.Object-oriented classification of tree species based on rule extraction from rough set[J].Remote Sensing Information,2013,28(4):90-97. (in Chinese)
[5]
BAATZMSCHÄPE A. Object-oriented and multi-scale image analysis in semantic networks2nd International Symposium: Operationalization of Remote Sensing. Enschede: ITC
1999713
BAATZM,SCHÄPE A. Object-oriented and multi-scale image analysis in semantic networks[C]//2nd International Symposium: Operationalization of Remote Sensing. Enschede: ITC,1999:7-13.
LIQ,GAO XZ,ZHANGT,et al.Optimal segmentation scale selection and evaluation for multi-layer image recognition and classification[J].Journal of Geo-information Science,2011,13(3):409-417. (in Chinese)
[7]
DR<inline-graphic href="0254-0037-43-5-710/img_7.jpg"/>GUŢL,EISANKC,STRASSERT.Local variance for multi-scale analysis in geomorphometry[J].Geomorphology,2011,130(3):162-172.
[8]
WOODCOCK CE,STRAHLER AH.The factor of scale in remote sensing[J].Remote sensing of Environment,1987,21(3):311-332.
[9]
马浩然.基于多层次分割的遥感影像面向对象森林分类[D].北京: 北京林业大学,2014.
MA HR.Object-based remote sensing image classification of forest based on multi-level segmentation[D].Beijng: Beijing Forestry University,2014. (in Chinese)
ZHANG YN,ZHANGC,ZHAO DL,et al.Object-oriented feature selection algorithms with ReliefFO and haiming genetic algorithm[J].Geomatics and Information Science of Wuhan University,2010,35(12):1444-1448. (in Chinese)
[11]
杨宁.高分辨率影像面向对象分类特征选择方法研究[D].西安: 西安科技大学,2012.
YANGN.Feature selection for object-oriented classification of high resolution remote sensing image[D].Xi’an: Xi’an University of Science and Technology,2012. (in Chinese)
[12]
HILLR,WILSONA,GEORGEM,et al.Mapping tree species in temperate deciduous woodland using time series multispectral data[J].Applied Vegetation Science,2010,13(1):86-99.
[13]
TIGGESJ,LAKEST,HOSTERTP.Urban vegetation classification: benefits of multitemporal RapidEye satellite data[J].Remote Sensing of Environment,2013,136:66-75.
LI CJ,LIU LY,WANG JH,et al.Comparison of two methods of fusing remote sensing images with fidelity of spectral information[J].Journal of Image and Graphics,2005,9(11):1376-1385. (in Chinese)
[15]
GMBH TG.eCognition user guide[M].München: Trappentreustr,2011.
LINH,LIUP,DU PJ,et al.Research on multi-scale segmentation algorithm of high-resolution remote sensing imagery based on improved statistical region merging[J].Computer Engineering and Applications,2012,48(18) :159-163. (in Chinese)
[17]
HARALICK RM,SHANMUGAMK,DINSTEIN IH.Textural features for image classification[J].IEEE Transactions on Systems, Man and Cybernetics,1973, (6):610-621.
[18]
李春干.面向对象的SPOT5图像森林分类研究[D].北京:北京林业大学,2009.
LI CG.Object-oriented classification of forest cover using SPOT5 imagery[D].Beijing: Beijing Forestry University,2009. (in Chinese)
YANGY.Comparative study on object-oriented classification technique with the World-View2 high-resolution image[D].Chengdu: Chengdu University of Technology,2015. (in Chinese)
ZHANGL,SHAO ZF,ZHOU XR,et al.Semi-supervised collaborative classification for hyperspectral remote sensing image with combination of cluster feature and SVM[J].Acta Geodaetica et Cartographica Sinica,2014,43(8):855-861. (in Chinese)
[21]
骆剑承,王钦敏,马江洪,等.遥感图像最大似然分类方法的 EM 改进算法[J].测绘学报,2002,31(3):234-239.
LUO JC,WANG QM,MA JH,et al.The EM-based maximum likelihood classifier for remotely sensed data[J].Acta Geodaetica et Cartographica Sinica,2002,31(3):234-239. (in Chinese)
MING DP,WANGQ,YANG JY.Spatial scale of remote sensing image and selection of optimal spatial resolution[J].Journal of Remote Sensing,2008,12(4):529-537. (in Chinese)
[23]
RICARDO D DS,PEDRINIH.Hyperspectral data classification improved by minimum spanning forests[J].Journal of Applied Remote Sensing,2016,10(2):1-17.
Table
3.
Feature selected with multi-temproal images
最终筛选的特征
Mean_627_NIR
Mean_423_NIR
Mean_317_NIR
GLCM_Mean_423_G
Length/Width
GLCM_Cont_627_NIR
GLCM_Homo_518_NIR
GLDV_Ent_518_R
GLDV_Ent_423_G
GLCM_Cor_423_NIR
NDVI_627
Stddev_317_NIR
GLCM_Ang_423_NIR
GLCM_Mean_627_NIR
GLDV_Mean_317_NIR
Std_dev_627_NIR
注:1)表中数字代表影像日期,R、G、NIR分别代表红、绿、近红外波段;2)表中纹理特征部分采用缩写,分别为同质度(homogeneity,Homo)、对比度(contrast,Cont)、角二阶矩(angular second moment,Ang)、,熵(entropy,Ent)、非相似度(dissimilarity,Dis)、均值(mean)、标准差(stddev)、相关(correlation,Cor).