基于面向对象的黄土丘陵沟壑区梯田信息提取研究
|
摘要
研究区位于黄土丘陵沟壑区第二副区的彭阳县,以国产GF-2影像为数据源,首先按地形复杂度生态分区,选取4个典型试验区;其次应用多尺度分割和光谱差异分割相结合的方法,对试验区影像进行分割;然后重点基于面向对象分类方法,实现梯田信息的自动提取;最后采用参考对象一致性误差(OCE)分割评价准则,以目视解译结果为参考对象,对梯田提取结果进行精度评价。结果表明,基于面向对象分类,不同生态分区遥感影像在最优分割阈值时,梯田信息自动提取精度均达到65%以上,结合后期处理,可以高效、准确地从国产GF-2影像中提取复杂地貌区的梯田及空间分布等相关信息,为黄土丘陵沟壑区农业生产规划、水土流失防治提供基础数据,也提供了新的思路和合理的技术方法。
In order to study the terrace extraction technology in loess hilly-gully region landscape,the study selected Pengyang County,which is located in the second sub region of Loess Hilly-gully Region Landscape,as the research area. The study chooses GF-2 remote sensing image as the basic data. According to the research of terrain complexity of ecological zones selected four typical experimental area. Firstly,the image of the experimental area is segmented by the combination of multi-scale segmentation and spectral difference segmentation. Then,the object-oriented classification method is used to extract the terrace information. Finally,the accuracy of the extraction results is evaluated by using the criterion of reference object consistency error( OCE) segmentation,and the results of visual interpretation are taken as the reference object. The results show that the accuracy of automatic extraction of terrace information is more than 65% based on the object-oriented classification method when the remote sensing images of different ecological regions are in the optimal segmentation threshold. Combined with the later processing,the area and spatial distribution information of terraced fields in complex geomorphic areas can be extracted efficiently and accurately from the GF-2 remote sensing image. It provides the basic data for agricultural production planning,prevention and control of soil erosion in loess hilly-gully region landscape.This method can be extended to other counties in the loess hilly-gully region landscape,and provides new ideas and reasonable technical methods for automatic extraction of terrace information in loess hilly-gully region landscape by using high resolution remote sensing images.
In order to study the terrace extraction technology in loess hilly-gully region landscape,the study selected Pengyang County,which is located in the second sub region of Loess Hilly-gully Region Landscape,as the research area. The study chooses GF-2 remote sensing image as the basic data. According to the research of terrain complexity of ecological zones selected four typical experimental area. Firstly,the image of the experimental area is segmented by the combination of multi-scale segmentation and spectral difference segmentation. Then,the object-oriented classification method is used to extract the terrace information. Finally,the accuracy of the extraction results is evaluated by using the criterion of reference object consistency error( OCE) segmentation,and the results of visual interpretation are taken as the reference object. The results show that the accuracy of automatic extraction of terrace information is more than 65% based on the object-oriented classification method when the remote sensing images of different ecological regions are in the optimal segmentation threshold. Combined with the later processing,the area and spatial distribution information of terraced fields in complex geomorphic areas can be extracted efficiently and accurately from the GF-2 remote sensing image. It provides the basic data for agricultural production planning,prevention and control of soil erosion in loess hilly-gully region landscape.This method can be extended to other counties in the loess hilly-gully region landscape,and provides new ideas and reasonable technical methods for automatic extraction of terrace information in loess hilly-gully region landscape by using high resolution remote sensing images.
引文
[1]张雨果,王飞,孙文义,等.基于面向对象的SPOT卫星影像梯田信息提取研究[J].水土保持研究,2016,23(6):345-351.
[2]赵俊峰,肖礼,黄懿梅,等.黄土丘陵区不同种植类型梯田2 m土层有机碳的分布特征[J].水土保持学报,2017,31(5):253-259.
[3]党恬敏,穆兴民,孙文义,等.高分辨率遥感影像梯田快速提取方法研究进展[J].人民黄河,2017,39(3):85-89,94.
[4]蔡宗磊,包妮沙,刘善军.国产高分一号数据估算草地植被覆盖度方法研究---以呼伦贝尔草原露天煤矿区为例[J].地理与地理信息科学,2017,33(2):32-38,44.
[5]孙小涛,周忠发,陈全,等.重点生态功能区水土流失敏感性评价与分布研究---以贵州省雷山县为例[J].水土保持学报,2016,30(6):73-78,133.
[6]蒲罗曼,张树文,王让虎,等.多源遥感影像的侵蚀沟信息提取分析[J].地理与地理信息科学,2016,32(1):90-94.
[7]于浩,刘志红,张晓萍,等.基于傅立叶变换的梯田纹理特征提取[J].国土资源遥感,2008(2):39-42.
[8]李斌兵,黄磊.基于面向对象技术的黄土丘陵沟壑区切沟遥感提取方法研究[J].水土保持研究,2013,20(3):115-119,124.
[9]张雨果.基于面向对象的遥感影像梯田信息提取研究[D].西安:西北农林科技大学,2016.
[10]陈明,周伟,袁涛.“GF-1”影像质量评价及矿区土地利用分类潜力研究[J].测绘科学技术学报,2015,32(5):494-499.
[11]石云.基于RS和GIS的退耕还林(草)动态监测与决策分析[D].西安:西北大学,2006.
[12]耿莎,石云.黄土丘陵区小流域景观格局与生态恢复措施和地形因子耦合关系研究[J].生态与农村环境学报,2015,31(2):197-203.
[13]何文秀,石云.黄土丘陵沟壑区地形复杂度分析[J].测绘科学,2015,40(10):146-152.
[14]杨存祥,朱建军,杨晓辉,等.彭阳县在推进水土保持生态文明建设中的实践与创新[J].中国水土保持,2013(12):70-72.
[15]孙攀,董玉森,陈伟涛,等.高分二号卫星影像融合及质量评价[J].国土资源遥感,2016,28(4):108-113.
[16]王荣,王昭生,刘晓曼.多尺度多准则的遥感影像线状地物信息提取[J].测绘科学,2016,41(11):146-150,165.
[17]熊华伟,俞春生,李小玉,等.基于高分辨率遥感影像的不透水面信息快速提取[J].国土与自然资源研究,2015(1):52-54.
[18]陈杰,陈铁桥,梅小明,等.基于最优尺度选择的高分辨率遥感影像丘陵农田提取[J].农业工程学报,2014,30(5):99-107.
[19]Polak M,Zhang H,Pi M H.An evaluation metric for image segmentation of multiple objects[J].Image and Vision Computing,2008,27(8):1 223-1 227.
[2]赵俊峰,肖礼,黄懿梅,等.黄土丘陵区不同种植类型梯田2 m土层有机碳的分布特征[J].水土保持学报,2017,31(5):253-259.
[3]党恬敏,穆兴民,孙文义,等.高分辨率遥感影像梯田快速提取方法研究进展[J].人民黄河,2017,39(3):85-89,94.
[4]蔡宗磊,包妮沙,刘善军.国产高分一号数据估算草地植被覆盖度方法研究---以呼伦贝尔草原露天煤矿区为例[J].地理与地理信息科学,2017,33(2):32-38,44.
[5]孙小涛,周忠发,陈全,等.重点生态功能区水土流失敏感性评价与分布研究---以贵州省雷山县为例[J].水土保持学报,2016,30(6):73-78,133.
[6]蒲罗曼,张树文,王让虎,等.多源遥感影像的侵蚀沟信息提取分析[J].地理与地理信息科学,2016,32(1):90-94.
[7]于浩,刘志红,张晓萍,等.基于傅立叶变换的梯田纹理特征提取[J].国土资源遥感,2008(2):39-42.
[8]李斌兵,黄磊.基于面向对象技术的黄土丘陵沟壑区切沟遥感提取方法研究[J].水土保持研究,2013,20(3):115-119,124.
[9]张雨果.基于面向对象的遥感影像梯田信息提取研究[D].西安:西北农林科技大学,2016.
[10]陈明,周伟,袁涛.“GF-1”影像质量评价及矿区土地利用分类潜力研究[J].测绘科学技术学报,2015,32(5):494-499.
[11]石云.基于RS和GIS的退耕还林(草)动态监测与决策分析[D].西安:西北大学,2006.
[12]耿莎,石云.黄土丘陵区小流域景观格局与生态恢复措施和地形因子耦合关系研究[J].生态与农村环境学报,2015,31(2):197-203.
[13]何文秀,石云.黄土丘陵沟壑区地形复杂度分析[J].测绘科学,2015,40(10):146-152.
[14]杨存祥,朱建军,杨晓辉,等.彭阳县在推进水土保持生态文明建设中的实践与创新[J].中国水土保持,2013(12):70-72.
[15]孙攀,董玉森,陈伟涛,等.高分二号卫星影像融合及质量评价[J].国土资源遥感,2016,28(4):108-113.
[16]王荣,王昭生,刘晓曼.多尺度多准则的遥感影像线状地物信息提取[J].测绘科学,2016,41(11):146-150,165.
[17]熊华伟,俞春生,李小玉,等.基于高分辨率遥感影像的不透水面信息快速提取[J].国土与自然资源研究,2015(1):52-54.
[18]陈杰,陈铁桥,梅小明,等.基于最优尺度选择的高分辨率遥感影像丘陵农田提取[J].农业工程学报,2014,30(5):99-107.
[19]Polak M,Zhang H,Pi M H.An evaluation metric for image segmentation of multiple objects[J].Image and Vision Computing,2008,27(8):1 223-1 227.