基于易康软件的QuickBird遥感影像林分类型识别——以福建省将乐林场为例
|
摘要
【目的】研究基于面向对象方法的林分类型识别,解决森林资源监测的核心问题。【方法】以福建省将乐林场为研究样本,采用基于Quick Bird遥感影像的蓝、绿、红、近红外4个多光谱波段为面向对象分类的试验数据,借助e Cognition Developer 8.7(易康)软件,设置10种分割尺度(25～250,步长为25),应用带有线性核函数支持向量机分类器(support vector machine,SVM),分别对每种分割尺度下的3组特征(单独光谱、光谱+纹理、光谱+纹理+空间)进行面向对象林分类型分类。【结果】以尺度参数150对Quick Bird遥感影像进行分割质量最高(ED3Modified=0.37)。10种尺度上,在光谱特征中加入纹理特征能够明显提高分类精度,但引入空间特征分类精度几乎无变化。基于光谱+纹理特征在分割尺度150时获得了最高分类精度(总精度达到85%,Kappa系数为0. 86)。【结论】分割尺度对面向对象林分类型识别精度有着重要影响。在所有尺度(25～250)下,光谱、纹理特征分类精度均高于单独使用光谱特征分类总精度,空间特征在林分类型分类中并没有起到作用。匹配良好的分割和参考对象时能够得到更高精度的分类结果,同时,轻微的过度分割或分割不足不会明显影响分类结果。基于易康软件的面向对象方法对Quick Bird多波段遥感数据进行林分类型分类能够获得比较满意的结果。
【Objective】Identification of forest stand is critical for forest resources monitoring.【Method】To study the extraction of forest stand information based on object oriented method,Quick Bird remote sensing image with multispectral bands( blue,green,red and near-infrared) was used as the experimental data,and 10 segmentation scales( 25-250,step size 25) were carried out using e Cognition Developer 8.7. For each segmentation scale,the support vector machine with linear kernel was applied to three combination features( spectrum,spectrum+ texture,spectrum+ texture+ space),respectively. 【Result】The results showed that segmentation scale was significant to forest stand identification,with a highest segmentation quality at segmentation scale of 150. At each of 10 segmentation scales,introducing texture features into spectral features could improve accuracy of classification; however,introducing spatial features into spectral features had no influence on accuracy of classification. So the highest accuracy of classification( OA = 85%; Kappa value is0. 86) was obtained based on the integration of spectral and texture features at segmentation scale of 150. 【Conclusion】Segmentation scale plays an important role in tree species classification. At all scales( 25-250),overall accuracy of spectral and texture features was higher than that of overall accuracy using spectral features alone. Spatial features did not play a role in forest classification. Matches between segmented and reference objects produced higher classification accurate,and slight over-and under-segmentations did not significantly affect the classifications. The object-based method based on e Cognition software can obtain satisfactory results for classification of stand types from Quick Bird multi-band remote sensing data.
【Objective】Identification of forest stand is critical for forest resources monitoring.【Method】To study the extraction of forest stand information based on object oriented method,Quick Bird remote sensing image with multispectral bands( blue,green,red and near-infrared) was used as the experimental data,and 10 segmentation scales( 25-250,step size 25) were carried out using e Cognition Developer 8.7. For each segmentation scale,the support vector machine with linear kernel was applied to three combination features( spectrum,spectrum+ texture,spectrum+ texture+ space),respectively. 【Result】The results showed that segmentation scale was significant to forest stand identification,with a highest segmentation quality at segmentation scale of 150. At each of 10 segmentation scales,introducing texture features into spectral features could improve accuracy of classification; however,introducing spatial features into spectral features had no influence on accuracy of classification. So the highest accuracy of classification( OA = 85%; Kappa value is0. 86) was obtained based on the integration of spectral and texture features at segmentation scale of 150. 【Conclusion】Segmentation scale plays an important role in tree species classification. At all scales( 25-250),overall accuracy of spectral and texture features was higher than that of overall accuracy using spectral features alone. Spatial features did not play a role in forest classification. Matches between segmented and reference objects produced higher classification accurate,and slight over-and under-segmentations did not significantly affect the classifications. The object-based method based on e Cognition software can obtain satisfactory results for classification of stand types from Quick Bird multi-band remote sensing data.
引文
[1]VIEIRA I C G,ALMEIDA A S D,DAVIDSON E A,et al. Classifying successional forests using landsat spectral properties and ecological characteristics in eastern Amaz8nia[J]. Remote Sensing of Environment,2003,87(4):470-481. DOI:10.1016/j.rse.2002.09.002.
[2] COLSTOUN E C B,STORY M H,THOMPSON C,et al.National park vegetation mapping using multi-temporal Landsat 7data and a decision tree classifier[J]. Remote Sensing of Environment,2003,85(3):316-327. DOI:10. 1016/S0034-4257(03)00010-5.
[3]WOLTER P T,MLADENOFF D J,HOST G E,et al. Improved forest classification in the northern Lake States using multitemporal landsat imagery[J]. Photogrammetric Engineering&Remote Sensing,1995,61(9):1129-1143. DOI:10. 1109/36.469496.
[4]CLARK M L,ROBERTS D A,CLARK D B. Hyperspectral discrimination of tropical rain forest tree species at leaf to crown scales[J]. Remote Sensing of Environment,2005,96(3):375-398. DOI:10.1016/j.rse.2005.03.009.
[5] GOODENOUGH D G, DYK A, NIEMANN K O, et al.Processing hyperion and a LI for forest classification[J]. IEEE Transactions on Geoscience&Remote Sensing,2003,41(6):1321-1331. DOI:10.1109/TGRS.2003. 813214.
[6]LAWRENCE R L,WOOD S D,SHELEY R L. Mapping invasive plants using hyperspectral imagery and Breiman Cutler classifications(random forest)[J]. Remote Sensing of Environment,2006,100(3):356-362.DOI:1 0.1016/j.rse.2005.10.014.
[7]IMMITZER M,ATZBERGER C,KOUKAL T. Tree species classification with random forest using very high spatial resolution 8-band WorldView-2 satellite data[J]. Remote Sensing,2002,4(9):2661-2693. DOI:10. 3390/rs4092661.
[8]SUN X Y,DU H Q,HAN N,et al. Synergistic use of landsat TM and SPOT5 imagery for object-based forest classification[J].Journal of Applied Remote Sensing,2014,8(1):801-807. DOI:10.1117/1.JRS.8.083550.
[9]JOHANSEN K,ARROYO L A,PHINN S,et al. Comparison of geo-object based and pixel-based change detection of riparian environments using high spatial resolution multi-spectral imagery[J]. Photogrammetric Engineering and Remote Sensing,2010,76(2):123-136.
[10]HOSSAIN S M Y,CASPERSEN J P. In-situ measurement of twig dieback and regrowth in mature Acer saccharum trees[J]. Forest Ecology&Management,2012,270(4):183-188. DOI:10.1016/j.foreco.2012.01.020.
[11]MUTANGA O,ADAM E,CHO M A. High density biomass estimation for wetland vegetation using World View-2 imagery and random forest regression algorithm[J]. International Journal of Applied Earth Observation and Geoinformation,2012,18(1):399-406. DOI:10.1016/j.jag.2012.03.012.
[12]GAULTON R,MALTHUS T J. Li DAR mapping of canopy gaps in continuous cover forests:a comparison of canopy height model and point cloud based techniques[J]. International Journal of Remote Sensing, 2008, 31(5):17-19. DOI:10.1080/01431160903380565.
[13] VEPAKOMMA U,ST-ONGE B,KNEESHAW D. Spatially explicit characterization of boreal forest gap dynamics using multitemporal lidar data[J]. Remote Sensing of Environment,2008,112(5):2326-2340. DOI:10. 1016/j.rse.2007.10.001.
[14]KIM M,MADDEN M,WARNER T. Estimation of optimal image object size for the segmentation of forest stands with multispectral IKONOS imagery[C]//Object-based Image Analysis. Berlin:Springer,2008:10-15.
[15]WANG L,SOUSA W P,GONG P. Integration of object-based and pixel-based classification for mapping mangroves with IKONOS imagery[J]. International Journal of Remote Sensing,2004,25(24):5655-5668. DOI:10.1080/014311602331291215.
[16]COMANICIU D,MEER P. Mean shift:a robust approach toward feature space analysis[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2002,24(5):603-619.DOI:10.1109/34.1000236.
[17]MICHEL J,YOUSSEFI D,GRIZONNET M. Stable mean-shift algorithm and its application to the segmentation of arbitrarily large remote sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing,2015,53(53):952-964. DOI:10.1109/TGRS.2014.2330857.
[18]BENZ U C,HOFMANN P,WILLHAUCK G,et al. Multi-resolution object-oriented fuzzy analysis of remote sensing data for GISready information[J]. International Journal of Photogrammetry&Remote Sensing,2004,58(3/4):239-258. DOI:10.1016/j. isprsjprs.2003.10.002.
[19]LI P J,GUO J C,SONG B Q,et al. A multilevel hierarchical image segmentation method for urban impervious surface mapping using very high resolution imagery[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2011,4(1):103-116. DOI:10.1109/JSTARS.2010.2074186.
[20]LI D,ZHANG G,WU Z,et al. An edge embedded marker-based watershed algorithm for high spatial resolution remote sensing image segmentation[J]. IEEE Transactions on Image Processing,2010,19(10):2781-2787. DOI:10.1109/TIP.2010.2049528.
[21] VINCENT L,SOILLE P. Watersheds in digital spaces:an efficient algorithm based on immersion simulations[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,1991,13(6):583-598.DOI:10.1109/34.87344.
[22]CHO M A,MATHIEU R,ASNER G P,et al. Mapping tree species composition in South African savannas using an integrated airborne spectral and Li DAR system[J]. Remote Sensing of Environment,2012,125(10):214-226. DOI:10. 1016/j. rse. 2012.07.010.
[23]MALAHLELA O,CHO M A,MUTANGA O. Mapping canopy gaps in an indigenous subtropical coastal forest using high-resolution World View-2 data[J]. International Journal of Remote Sensing,2014,3517(17):6397-6417. DOI:10. 1080/01431161.2014.954061.
[24]IM J,JENSEN J R,HODGSON M E. Object-based land cover classification using high-posting-density LIDAR data[J].Mapping Science and Remote Sensing,2008,45(2):209-229.DOI:10.2747/1548-1603.45.2.209.
[25]IM J,JENSEN J R,TULLIS J A. Object-based change detection using correlation image analysis and image segmentation[J]. International Journal of Remote Sensing,2007,29(2):399-423.DOI:10.1080/0143116 0601075582.
[26]YANG J,HE Y H,WENG Q H. An automated method to parameterize segmentation scale by enhancing intrasegment homogeneity and intersegment heterogeneity[J]. IEEE Geoscience&Remote Sensing Letters,2015,12(6):1282-1286. DOI:10. 1109/LGRS.2015.2393255.
[27] YOUNG T Y,FU K S. Handbook of pattern recognition and image processing[M].New York:Academic Press,1986:23-28.
[28] ECOGNITION B. Definiens imaging gmb H[M]. Sunnyvale:Trimble,2010:34-40.
[29]GITELSON A A,KAUFMAN Y J,MERZLYAK M N. Use of a green channel in remote sensing of global vegetation from EOSMODIS[J]. Remote Sensing of Environment,1996,58(3):289-298.DOI:10.1016/s0034-4257(96)00072-1.
[30] VAPNIK V N. The nature of statistical learning theory[M].Berlin:Springer Verlag,2000:67-72.
[31]CLINTON N,HOLT A,SCARBOROUGH J,et al. Accuracy assessment measures for object-based image segmentation goodness[J]. Photogrammetric Engineering and Remote Sensing,2010,76(3):289-299.
[32]LIU Y,BIAN L,MENG Y,et al. Discrepancy measures for selecting optimal combination of parameter values in object-based image analysis[J]. Journal of Photogrammetry and Remote Sensing,2012,68(1):144-156. DOI:10. 10106/j. isprsjprs. 2012.01.007.
[33] MOLLER M,LYMBURNER L,VOLK M. The comparison index:a tool for assessing the accuracy of image segmentation[J]. International Journal of Applied Earth Observation&Geoinformation,2007,9(3):311-321. DOI:10. 1016/j. jag. 2006.10.002.
[34]ZHAN Q M,MOLENAAR M,TEMPFLI K,et al. Quality assessment for geo-spatial objects derived from remotely sensed data[J]. International Journal of Remote Sensing,2007,26(14):2953-2974.. DOI:10.1080/01431 160500057764.
[35]JANSSEN L L F,WEI F J M. Accuracy assessment of satellite derived land-cover data:a review[J]. Photogrammetric Engineering&Remote Sensing,1994,60(4):419-426.
[36]KE Y H,QUACKENBUSH L J,IM J. Synergistic use of QuickBird multispectral imagery and LIDAR data for object-based forest species classification[J]. Remote Sensing of Environment,2010,114(6):1141-1154. DOI:10.1016/j.rse.2010.01.002.
[37]WANG L,SOUSA W P,GONG P. Integration of object-based and pixel-based classification for mapping mangroves with IKONOS imagery[J]. International Journal of Remote Sensing,2004,25(24):5655-5668. DOI:10.1080/014311602331291215.
[38]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. DOI:10.13332/j.1000-1522.20140311.LIU H P,AN H J,WANG B,et al. Tree species classification using WorldView-2 images based on recursive texture feature elimination[J].Journal of Beijing Forestry University,2015,37(8):53-59.
[39]王妮,彭世揆,李明诗.基于树种分类的高分辨率遥感数据纹理特征分析[J].浙江农林大学学报,2012,29(2):210-217.DOI:10.3969/j.issn.2095-0756.2012.02.010.WANG N,PENG S K,LI M S. High-resolution remote sensing of textural images for tree species classification[J]. Journal of Zhejiang A&F University,2012,29(2):210-217.
[40]YU Q,GONG P,CLINTON N,et al. Object-based detailed vegetation classification with airborne high spatial resolution remote sensing imagery[J]. Photogrammetric Engineering&Remote Sensing,2006,72(7):799-811.
[2] COLSTOUN E C B,STORY M H,THOMPSON C,et al.National park vegetation mapping using multi-temporal Landsat 7data and a decision tree classifier[J]. Remote Sensing of Environment,2003,85(3):316-327. DOI:10. 1016/S0034-4257(03)00010-5.
[3]WOLTER P T,MLADENOFF D J,HOST G E,et al. Improved forest classification in the northern Lake States using multitemporal landsat imagery[J]. Photogrammetric Engineering&Remote Sensing,1995,61(9):1129-1143. DOI:10. 1109/36.469496.
[4]CLARK M L,ROBERTS D A,CLARK D B. Hyperspectral discrimination of tropical rain forest tree species at leaf to crown scales[J]. Remote Sensing of Environment,2005,96(3):375-398. DOI:10.1016/j.rse.2005.03.009.
[5] GOODENOUGH D G, DYK A, NIEMANN K O, et al.Processing hyperion and a LI for forest classification[J]. IEEE Transactions on Geoscience&Remote Sensing,2003,41(6):1321-1331. DOI:10.1109/TGRS.2003. 813214.
[6]LAWRENCE R L,WOOD S D,SHELEY R L. Mapping invasive plants using hyperspectral imagery and Breiman Cutler classifications(random forest)[J]. Remote Sensing of Environment,2006,100(3):356-362.DOI:1 0.1016/j.rse.2005.10.014.
[7]IMMITZER M,ATZBERGER C,KOUKAL T. Tree species classification with random forest using very high spatial resolution 8-band WorldView-2 satellite data[J]. Remote Sensing,2002,4(9):2661-2693. DOI:10. 3390/rs4092661.
[8]SUN X Y,DU H Q,HAN N,et al. Synergistic use of landsat TM and SPOT5 imagery for object-based forest classification[J].Journal of Applied Remote Sensing,2014,8(1):801-807. DOI:10.1117/1.JRS.8.083550.
[9]JOHANSEN K,ARROYO L A,PHINN S,et al. Comparison of geo-object based and pixel-based change detection of riparian environments using high spatial resolution multi-spectral imagery[J]. Photogrammetric Engineering and Remote Sensing,2010,76(2):123-136.
[10]HOSSAIN S M Y,CASPERSEN J P. In-situ measurement of twig dieback and regrowth in mature Acer saccharum trees[J]. Forest Ecology&Management,2012,270(4):183-188. DOI:10.1016/j.foreco.2012.01.020.
[11]MUTANGA O,ADAM E,CHO M A. High density biomass estimation for wetland vegetation using World View-2 imagery and random forest regression algorithm[J]. International Journal of Applied Earth Observation and Geoinformation,2012,18(1):399-406. DOI:10.1016/j.jag.2012.03.012.
[12]GAULTON R,MALTHUS T J. Li DAR mapping of canopy gaps in continuous cover forests:a comparison of canopy height model and point cloud based techniques[J]. International Journal of Remote Sensing, 2008, 31(5):17-19. DOI:10.1080/01431160903380565.
[13] VEPAKOMMA U,ST-ONGE B,KNEESHAW D. Spatially explicit characterization of boreal forest gap dynamics using multitemporal lidar data[J]. Remote Sensing of Environment,2008,112(5):2326-2340. DOI:10. 1016/j.rse.2007.10.001.
[14]KIM M,MADDEN M,WARNER T. Estimation of optimal image object size for the segmentation of forest stands with multispectral IKONOS imagery[C]//Object-based Image Analysis. Berlin:Springer,2008:10-15.
[15]WANG L,SOUSA W P,GONG P. Integration of object-based and pixel-based classification for mapping mangroves with IKONOS imagery[J]. International Journal of Remote Sensing,2004,25(24):5655-5668. DOI:10.1080/014311602331291215.
[16]COMANICIU D,MEER P. Mean shift:a robust approach toward feature space analysis[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2002,24(5):603-619.DOI:10.1109/34.1000236.
[17]MICHEL J,YOUSSEFI D,GRIZONNET M. Stable mean-shift algorithm and its application to the segmentation of arbitrarily large remote sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing,2015,53(53):952-964. DOI:10.1109/TGRS.2014.2330857.
[18]BENZ U C,HOFMANN P,WILLHAUCK G,et al. Multi-resolution object-oriented fuzzy analysis of remote sensing data for GISready information[J]. International Journal of Photogrammetry&Remote Sensing,2004,58(3/4):239-258. DOI:10.1016/j. isprsjprs.2003.10.002.
[19]LI P J,GUO J C,SONG B Q,et al. A multilevel hierarchical image segmentation method for urban impervious surface mapping using very high resolution imagery[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2011,4(1):103-116. DOI:10.1109/JSTARS.2010.2074186.
[20]LI D,ZHANG G,WU Z,et al. An edge embedded marker-based watershed algorithm for high spatial resolution remote sensing image segmentation[J]. IEEE Transactions on Image Processing,2010,19(10):2781-2787. DOI:10.1109/TIP.2010.2049528.
[21] VINCENT L,SOILLE P. Watersheds in digital spaces:an efficient algorithm based on immersion simulations[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,1991,13(6):583-598.DOI:10.1109/34.87344.
[22]CHO M A,MATHIEU R,ASNER G P,et al. Mapping tree species composition in South African savannas using an integrated airborne spectral and Li DAR system[J]. Remote Sensing of Environment,2012,125(10):214-226. DOI:10. 1016/j. rse. 2012.07.010.
[23]MALAHLELA O,CHO M A,MUTANGA O. Mapping canopy gaps in an indigenous subtropical coastal forest using high-resolution World View-2 data[J]. International Journal of Remote Sensing,2014,3517(17):6397-6417. DOI:10. 1080/01431161.2014.954061.
[24]IM J,JENSEN J R,HODGSON M E. Object-based land cover classification using high-posting-density LIDAR data[J].Mapping Science and Remote Sensing,2008,45(2):209-229.DOI:10.2747/1548-1603.45.2.209.
[25]IM J,JENSEN J R,TULLIS J A. Object-based change detection using correlation image analysis and image segmentation[J]. International Journal of Remote Sensing,2007,29(2):399-423.DOI:10.1080/0143116 0601075582.
[26]YANG J,HE Y H,WENG Q H. An automated method to parameterize segmentation scale by enhancing intrasegment homogeneity and intersegment heterogeneity[J]. IEEE Geoscience&Remote Sensing Letters,2015,12(6):1282-1286. DOI:10. 1109/LGRS.2015.2393255.
[27] YOUNG T Y,FU K S. Handbook of pattern recognition and image processing[M].New York:Academic Press,1986:23-28.
[28] ECOGNITION B. Definiens imaging gmb H[M]. Sunnyvale:Trimble,2010:34-40.
[29]GITELSON A A,KAUFMAN Y J,MERZLYAK M N. Use of a green channel in remote sensing of global vegetation from EOSMODIS[J]. Remote Sensing of Environment,1996,58(3):289-298.DOI:10.1016/s0034-4257(96)00072-1.
[30] VAPNIK V N. The nature of statistical learning theory[M].Berlin:Springer Verlag,2000:67-72.
[31]CLINTON N,HOLT A,SCARBOROUGH J,et al. Accuracy assessment measures for object-based image segmentation goodness[J]. Photogrammetric Engineering and Remote Sensing,2010,76(3):289-299.
[32]LIU Y,BIAN L,MENG Y,et al. Discrepancy measures for selecting optimal combination of parameter values in object-based image analysis[J]. Journal of Photogrammetry and Remote Sensing,2012,68(1):144-156. DOI:10. 10106/j. isprsjprs. 2012.01.007.
[33] MOLLER M,LYMBURNER L,VOLK M. The comparison index:a tool for assessing the accuracy of image segmentation[J]. International Journal of Applied Earth Observation&Geoinformation,2007,9(3):311-321. DOI:10. 1016/j. jag. 2006.10.002.
[34]ZHAN Q M,MOLENAAR M,TEMPFLI K,et al. Quality assessment for geo-spatial objects derived from remotely sensed data[J]. International Journal of Remote Sensing,2007,26(14):2953-2974.. DOI:10.1080/01431 160500057764.
[35]JANSSEN L L F,WEI F J M. Accuracy assessment of satellite derived land-cover data:a review[J]. Photogrammetric Engineering&Remote Sensing,1994,60(4):419-426.
[36]KE Y H,QUACKENBUSH L J,IM J. Synergistic use of QuickBird multispectral imagery and LIDAR data for object-based forest species classification[J]. Remote Sensing of Environment,2010,114(6):1141-1154. DOI:10.1016/j.rse.2010.01.002.
[37]WANG L,SOUSA W P,GONG P. Integration of object-based and pixel-based classification for mapping mangroves with IKONOS imagery[J]. International Journal of Remote Sensing,2004,25(24):5655-5668. DOI:10.1080/014311602331291215.
[38]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. DOI:10.13332/j.1000-1522.20140311.LIU H P,AN H J,WANG B,et al. Tree species classification using WorldView-2 images based on recursive texture feature elimination[J].Journal of Beijing Forestry University,2015,37(8):53-59.
[39]王妮,彭世揆,李明诗.基于树种分类的高分辨率遥感数据纹理特征分析[J].浙江农林大学学报,2012,29(2):210-217.DOI:10.3969/j.issn.2095-0756.2012.02.010.WANG N,PENG S K,LI M S. High-resolution remote sensing of textural images for tree species classification[J]. Journal of Zhejiang A&F University,2012,29(2):210-217.
[40]YU Q,GONG P,CLINTON N,et al. Object-based detailed vegetation classification with airborne high spatial resolution remote sensing imagery[J]. Photogrammetric Engineering&Remote Sensing,2006,72(7):799-811.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |