高级检索

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于WorldView2和GF-2的面向对象多指标综合植被变化分析

朱方嫣 沈文娟 李明诗

朱方嫣, 沈文娟, 李明诗. 基于WorldView2和GF-2的面向对象多指标综合植被变化分析[J]. 北京林业大学学报, 2019, 41(11): 54-65. doi: 10.13332/j.1000-1522.20180435
引用本文: 朱方嫣, 沈文娟, 李明诗. 基于WorldView2和GF-2的面向对象多指标综合植被变化分析[J]. 北京林业大学学报, 2019, 41(11): 54-65. doi: 10.13332/j.1000-1522.20180435
Zhu Fangyan, Shen Wenjuan, Li Mingshi. Object-oriented multi-index integrated vegetation change analysis based on WorldView2 and GF-2[J]. Journal of Beijing Forestry University, 2019, 41(11): 54-65. doi: 10.13332/j.1000-1522.20180435
Citation: Zhu Fangyan, Shen Wenjuan, Li Mingshi. Object-oriented multi-index integrated vegetation change analysis based on WorldView2 and GF-2[J]. Journal of Beijing Forestry University, 2019, 41(11): 54-65. doi: 10.13332/j.1000-1522.20180435

基于WorldView2和GF-2的面向对象多指标综合植被变化分析

doi: 10.13332/j.1000-1522.20180435
基金项目: 国家自然科学基金项目(31670552),江苏省高校优势学科建设项目(PAPD),江苏省青蓝工程项目(2017),江苏省高校研究生科研创新项目(KYLX15_0908),中国博士后科学基金资助项目
详细信息
    作者简介:

    朱方嫣。主要研究方向:遥感与地理信息系统。Email:15720613089@163.com  地址:210037江苏省南京市玄武区龙蟠路159号南京林业大学林学院

    责任作者:

    李明诗,博士,教授。主要研究方向:遥感与地理信息系统。Email:nfulms@njfu.edu.cn  地址:同上

  • 中图分类号: S771.8;TP79

Object-oriented multi-index integrated vegetation change analysis based on WorldView2 and GF-2

  • 摘要: 目的利用高分辨率卫星影像获取精确的植被变化信息对植被资源合理利用及可持续经营有重要意义。传统的基于像元的直接变化检测法容易产生椒盐噪声,而用面向对象分类法结果又严重依赖于分类精度。本文在分析现有算法优劣势基础上,力图找到一种针对高分辨率遥感数据进行植被变化检测的相对客观算法,并验证其有效性。方法基于现有的多指标综合变化分析算法(MIICA),提出了面向对象的MIICA。本算法用准确率(P)和查全率(R)分析确定的最优分割参数对前后两期跨传感器影像进行统一分割,利用分割获得的对象影像进行特征参数提取,并用ROC曲线法选择合适的阈值进行变化信息提取并整合,最终获得植被变化位置及方向(植被增多或减少)。结果经与基于像元的MIICA及面向对象分类法的比较,本方法的生产者精度高于基于像元的MIICA,用户精度高于面向对象分类法,并且总体精度和Kappa系数分别达到了0.880和0.805。本方法能更好地反映植被变化的位置及形状,也能较准确地检测出一些面积微小的变化。结论面向对象的MIICA能弥补基于像元的MIICA和面向对象分类的缺点,提高检测精度,对存在高人为影响的森林公园或自然保护区植被变化分析、植被资源合理利用及可持续经营有重要意义。

     

  • 图  1  研究区位置

       右图为Worldview2波段4、3、2的假彩色合成图像。The right figure is a false color composite using Worldview 2 band 4, 3 and 2.

    Figure  1.  Location of the study area

    图  2  面向对象的MIICA流程图

    dNDVI. 差异归一化植被指数;CV. 变化向量;RCVMAX. 相对变化向量最大值。 dNDVI, differenced normalized difference vegetation index; CV, change vector; RCVMAX, relative change vector maximum.

    Figure  2.  Flowchart of the object-oriented MIICA

    图  3  P值随合并尺度和分割尺度的变化

    a. 当分割尺度为0时P值随合并尺度的变化;b. 当合并尺度为75时P值随分割尺度的变化。 a, trends of P with the change of merging scale when the segmentation scale was 0; b, trends of P with the change of segmentation scale when the merging scale was 75.

    Figure  3.  Trends of P with the change of merging scale and segmentation scale

    图  4  影像分割结果

    a. 2015年WorldView2影像;b. WorldView2影像分割结果;c. 2018年GF-2影像;d. GF-2影像分割结果。 a, WorldView2 image; b, WorldView2 image segmentation result; c, 2018 GF-2 image; d, GF-2 image segmentation result.

    Figure  4.  Segmentation results of images

    图  5  面向对象的特征参数影像

    a. 面向对象的dNDVI影像;b. 面向对象的CV影像;c. 面向对象的RCVMAX影像。a, object-oriented dNDVI image; b, object-oriented CV image; c, object-oriented RCVMAX image.

    Figure  5.  Object-oriented feature parameter images

    图  6  CV在不同阈值下的ROC曲线

    Figure  6.  ROC curves of CV at different thresholds

    图  7  特征参数在不同变化阈值下FPR和TPR的ROC曲线

    a. dNDVI植被减少阈值;b. dNDVI植被增多阈值;c. RCVMAX正值部分阈值;d. RCVMAX负值部分阈值。 a, dNDVI thresholds of vegetation loss; b, dNDVI thresholds of vegetation gain; c, RCVMAX thresholds of positive part; d, RCVMAX thresholds of negative part.

    Figure  7.  ROC curves of FPR and TPR for characteristic parameters at different change thresholds

    图  8  紫金山2015—2018年植被变化专题图

    Figure  8.  The matic map of vegetation change in the Purple Mountains from 2015−2018

    图  9  不同方法检测结果图

    a. 2015年影像;b. 2018年影像;c. 基于像元的MIICA结果;d. 面向对象的MIICA结果;e. 面向对象分类法结果。①、②分别指示不同图像的相同位置。 a, image in 2015; b, image in 2018; c, pixel-based MIICA result; d, object-oriented MIICA result; e, object-oriented classification result. ① and ② indicate the same positions in different images, respectively.

    Figure  9.  Change detection results of different methods

    表  1  不同CV阈值下的FPR和TPR

    Table  1.   FPR and TPR at different CV thresholds

    正类率 Positive rateCV 阈值 CV threshold
    01002003003504005006007008009001 000
    FPR 1.00 0.64 0.38 0.21 0.19 0.13 0.08 0.06 0.03 0.02 0.01 0.00
    TPR 1.00 0.94 0.89 0.84 0.82 0.77 0.69 0.65 0.51 0.44 0.37 0.00
    注:FPR为假正类率,TPR为真正类率。Notes: FPR is false positive rate,TPR is true positive rate.
    下载: 导出CSV

    表  2  不同dNDVI阈值下的FPR和TPR(植被减少)

    Table  2.   FPR and TPR at different dNDVI thresholds (vegetation loss)

    正类率 Positive ratedNDVI 阈值(植被减少) dNDVI threshold (vegetation loss)
    0.000.110.120.130.140.150.160.200.501.00
    FPR 1.00 0.69 0.56 0.42 0.22 0.18 0.17 0.08 0.00 0.00
    TPR 1.00 0.98 0.97 0.95 0.87 0.84 0.83 0.65 0.13 0.00
    下载: 导出CSV

    表  6  本研究提取植被变化所用的参数阈值

    Table  6.   Parametric thresholds used to extract vegetation change in this study

    项目
    Item
    植被减少
    Vegetation loss
    植被增多
    Vegetation gain
    dNDVI > 0.16 < − 0.21
    CV > 350 > 350
    RCVMAX < − 0.006或 > 0.006 < − 0.006或 > 0.006
    下载: 导出CSV

    表  3  不同dNDVI阈值下的FPR和TPR(植被增多)

    Table  3.   FPR and TPR at different dNDVI thresholds (vegetation gain)

    正类率 Positive ratedNDVI 阈值(植被增多) dNDVI threshold (vegetation gain)
    0.00− 0.10− 0.15− 0.20− 0.21− 0.22− 0.25− 0.30− 0.40− 0.50
    FPR 1.00 0.56 0.31 0.26 0.13 0.08 0.06 0.04 0.01 0.00
    TPR 1.00 0.99 0.96 0.95 0.87 0.82 0.78 0.71 0.31 0.00
    下载: 导出CSV

    表  4  不同RCVMAX阈值下的FPR和TPR(正值部分)

    Table  4.   FPR and TPR at different RCVMAX thresholds (positive part)

    正类率 Positive rateRCVMAX(阈值正值部分) RCVMAX threshold (positive part)
    0.00 0.001 0.002 0.004 0.006 0.0080.010.020.040.06
    FPR 1.00 0.45 0.28 0.23 0.16 0.08 0.06 0.04 0.02 0.00
    TPR 1.00 0.94 0.91 0.89 0.84 0.72 0.67 0.59 0.41 0.00
    下载: 导出CSV

    表  5  不同RCVMAX阈值下的FPR和TPR(负值部分)

    Table  5.   FPR and TPR at different RCVMAX thresholds (negative part)

    正类率 Positive rateRCVMAX阈值(负值部分) RCVMAX thresholds (negative part)
    0.00− 0.001− 0.002− 0.003− 0.004− 0.005− 0.006− 0.01− 0.02− 0.03
    FPR 1.00 0.83 0.63 0.55 0.38 0.31 0.24 0.07 0.00 0.00
    TPR 1.00 0.97 0.93 0.91 0.83 0.79 0.74 0.47 0.02 0.00
    下载: 导出CSV

    表  7  紫金山2015—2018年植被变化面积及比例

    Table  7.   Vegetation change area and rate in the Purple Mountains during 2015−2018

    植被变化状况 Vegetation change面积 Area/m2比例 Rate/%
    植被减少 Vegetation loss 356 102 1.21
    植被增多 Vegetation gain 268 901 0.92
    下载: 导出CSV

    表  8  不同检测方法精度的对比

    Table  8.   Comparison of accuracies from different change detection methods

    精度相关参数
    Parameters of precision
    基于像元的MIICA
    Pixel-based MIICA
    面向对象的MIICA
    Object-oriented MIICA
    面向对象分类法
    Object-oriented classification
    植被减少Vegetation loss植被增多 Vegetation gain 植被减少 Vegetation loss植被增多 Vegetation gain 植被减少 Vegetation loss植被增多 Vegetation gain
    用户精度 User’s accuracy 0.927 0.957 0.896 0.944 0.712 0.780
    生产者精度 Producer’s accuracy 0.717 0.750 0.811 0.850 0.887 0.883
    总体精度 Overall accuracy 0.844 0.880 0.791
    Kappa 系数 Kappa coefficient 0.739 0.805 0.678
    下载: 导出CSV
  • [1] 王琰, 舒宁, 龚龑. 高分辨率遥感影像土地利用变化检测方法研究[J]. 国土资源遥感, 2012, 24(1):43−47. doi: 10.6046/gtzyyg.2012.01.08

    Wang Y, Shu N, Gong Y. A study of land use change detection based on high resolution remote sensing images[J]. Remote Sensing for Land & resources, 2012, 24(1): 43−47. doi: 10.6046/gtzyyg.2012.01.08
    [2] Deng J S, Wang K, Deng Y H, et al. PCA-based land-use change detection and analysis using multitemporal and multisensor satellite data[J]. International Journal of Remote Sensing, 2008, 29(16): 4823−4838. doi: 10.1080/01431160801950162
    [3] 赖祖龙, 申邵洪, 程新文, 等. 基于图斑的高分辨率遥感影像变化检测[J]. 测绘通报, 2009, 16(3):16−19.

    Lai Z L, Shen S H, Cheng X W, et al. Change detection of high-resolution remote sensing image based on map spot[J]. Bulletin of Surveying and Mapping, 2009, 16(3): 16−19.
    [4] Weismiller R A, Kristof S J, Scholz D K, et al. Change detection in coastal zone environments[J]. Photogrammetric Engineering & Remote Sensing, 1997, 43(12): 1533−1539.
    [5] 杨存建, 欧晓昆, 党承林, 等. 森林植被动态变化信息的遥感检测[J]. 地球信息科学学报, 2000, 2(4):71−74. doi: 10.3969/j.issn.1560-8999.2000.04.016

    Yang C J, Ou X K, Dang C L, et al. Detecting the change information of forest vegetation in Lushui County of Yunnan Province[J]. Geo-Information Science, 2000, 2(4): 71−74. doi: 10.3969/j.issn.1560-8999.2000.04.016
    [6] Chen J, Lu M, Chen X, et al. A spectral gradient difference based approach for land cover change detection[J]. Isprs Journal of Photogrammetry & Remote Sensing, 2013, 85(2): 1−12.
    [7] Jin S M, Yang L M, Danielson P, et al. A comprehensive change detection method for updating the National Land Cover Database to circa 2011[J]. Remote Sensing of Environment, 2013, 132(10): 159−175.
    [8] 韩飞. 光谱梯度差分与面向对象方法相结合的高分辨率遥感影像变化检测[D]. 成都: 西南交通大学, 2016.

    Han F. The change detection combining spectral gradient difference with object-oriented method for high resolution remote sensing images[D]. Chengdu: Southwest Jiaotong University, 2016.
    [9] Baatz M, Schäpe A. Object-oriented and multi-scale image analysis in semantic networks[C]//Proceeding of 2nd International Symposium: Operationalization of Remote Sensing, Enschede: ITC, 1999: 87−109.
    [10] Chen G, Hay G J, Carvalho L M T, et al. Object-based change detection[J]. International Journal of Remote Sensing, 2012, 33(14): 4434−4457. doi: 10.1080/01431161.2011.648285
    [11] Howell R G, Petersen S L. A comparison of change detection measurements using object-based and pixel-based classification methods on western juniper dominated woodlands in eastern Oregon[J]. Aims Environmental Science, 2017, 4(2): 348−357. doi: 10.3934/environsci.2017.2.348
    [12] Gamanya R, Maeyer P D, Dapper M D. Object-oriented change detection for the city of Harare, Zimbabwe[J]. Expert Systems with Applications, 2009, 36(1): 571−588. doi: 10.1016/j.eswa.2007.09.067
    [13] Desclée B, Bogaert P, Defourny P. Forest change detection by statistical object-based method[J]. Remote Sensing of Environment, 2006, 102(1): 1−11.
    [14] 李亮, 王蕾, 孙晓鹏, 等. 面向对象变化向量分析的遥感影像变化检测[J]. 遥感信息, 2017, 32(6):71−77. doi: 10.3969/j.issn.1000-3177.2017.06.012

    Li L, Wang L, Sun X P, et al. Remote sensing change detection method based on object-oriented change vector analysis[J]. Remote Sensing Information, 2017, 32(6): 71−77. doi: 10.3969/j.issn.1000-3177.2017.06.012
    [15] 刘璐. 南京紫金山植被变化及其对环境质量的影响[J]. 福建农业, 2015(6):209−211.

    Liu L. Changes of vegetation and its effect on environmental quality in Purple Mountain, Nanjing[J]. Fujian Agric, 2015(6): 209−211.
    [16] 董丽娜, 徐海兵, 居峰, 等. 南京紫金山国家森林公园植物多样性现状及保护对策[J]. 江苏林业科技, 2011, 38(1):30−35. doi: 10.3969/j.issn.1001-7380.2011.01.008

    Dong L N, Xu H B, Ju F, et al. Plant diversity and its conservation strategies in Zijin Mountain National Forest Park in Nanjing[J]. Journal of Jiangsu Forestry Science & Technology, 2011, 38(1): 30−35. doi: 10.3969/j.issn.1001-7380.2011.01.008
    [17] Qi C, Lalasia B M. Assessment of Vegetation Response to Ungulate Removal Utilizing High Resolution Remotely Sensed Data[R/OL]. Status Report for the Makua and Oahu Implementation Plans, 2014 [2018−09−16]. http://manoa.hawaii.edu/hpicesu/DPW/2014_YER/ES5.pdf.
    [18] 胡茂莹. 基于高分二号遥感影像面向对象的城市房屋信息提取方法研究[D]. 长春: 吉林大学, 2016.

    Hu M Y. Research on the object-oriented method of urban building extraction with GF-2 remote-sensing imagery[D]. Changchun: Jilin University, 2016.
    [19] 李明诗, 梅昭容. 土地覆盖变化检测中不同相对辐射归一化方法的评价[J]. 南京林业大学学报(自然科学版), 2017, 41(4):108−114.

    Li M S, Mei Z R. Performance assessment of different relative radiometric normalization approaches applied to land cover change detection[J]. Journal of Nanjing Forestry University (Natural Sciences Edition), 2017, 41(4): 108−114.
    [20] 彭士纯, 柳健. 基于高通滤波的多光谱图像融合方法[J]. 计算机应用研究, 2007, 24(8):218−219. doi: 10.3969/j.issn.1001-3695.2007.08.070

    Peng S C, Liu J. Multi-spectral image fusion method based on high pass filter[J]. Application Research of Computers, 2007, 24(8): 218−219. doi: 10.3969/j.issn.1001-3695.2007.08.070
    [21] Witharana C, Larue M A, Lynch H J. Benchmarking of data fusion algorithms in support of earth observation based Antarctic wildlife monitoring[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2016, 113: 124−143. doi: 10.1016/j.isprsjprs.2015.12.009
    [22] 龚建周, 刘彦随, 夏北成, 等. 小波基及其参数对遥感影像融合图像质量的影响[J]. 地理与地理信息科学, 2010, 26(2):6−10.

    Gong J Z, Liu Y S, Xia B C, et al. Effect of wavelet basis and decomposition levels on performance of fusion images from remotely sensed data[J]. Geography and Geo-Information Science, 2010, 26(2): 6−10.
    [23] 刘春燕. 图像分割评价方法研究[D]. 西安: 西安电子科技大学, 2011.

    Liu C Y. Survey on evaluation methods of image segmentation algorithms[D]. Xi’an: Xidian University, 2011.
    [24] 朱成杰, 杨世植, 崔生成, 等. 面向对象的高分辨率遥感影像分割精度评价方法[J/OL]. 强激光与粒子束, 2015, 27(6): 27061007[2018−06−16]. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=qjgylzs201506007.

    Zhu C J, Yang S Z, Cui S C, et al. Accuracy evaluating method for object-based segmentation of high resolution remote sensing image[J/OL]. High Power Laser and Particle Beams, 2015, 27(6): 27061007 [2018−06−16]. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=qjgylzs201506007.
    [25] LópezGarcía M J, Caselles V. Mapping burns and natural reforestation using thematic Mapper data[J]. Geocarto International, 1991, 6(1): 31−37.
    [26] Hanley J A. The meaning and use of the areas under a receiver operating characteristic (ROC) curve[J]. Radiology, 1982, 143(1): 29−36. doi: 10.1148/radiology.143.1.7063747
    [27] Leshowitz B. Comparison of ROC curves from one- and two-interval rating-scale procedures[J]. The Journal of the Acoustical Society of America, 1969, 46(2B): 399−402. doi: 10.1121/1.1911703
    [28] Banko G. A review of assessing the accuracy of classifications of remotely sensed data and of methods including remote sensing data in forest inventory[R]. Vienna: International Institute for Applied Systems Analysis Interim Report, 1998: IR-98-081.
    [29] Congalton R G. A review of assessing the accuracy of classifications of remotely sensed data[J]. Remote Sensing of Environment, 1991, 37(1): 35−46.
  • 加载中
图(9) / 表(8)
计量
  • 文章访问数:  948
  • HTML全文浏览量:  493
  • PDF下载量:  53
  • 被引次数: 0
出版历程
  • 收稿日期:  2018-12-25
  • 修回日期:  2019-06-23
  • 网络出版日期:  2019-09-29
  • 刊出日期:  2019-11-01

目录

    /

    返回文章
    返回