参数优化随机森林算法的土地覆盖分类
|
摘要
针对随机森林算法进行土地覆盖分类时无法确定参数组合以得到最优分类结果的问题,该文提出了两种随机森林算法的参数优化方法。以北京市昌平区为研究区,应用Landsat TM影像,实现了基于光谱值、纹理特征和专题特征的随机森林土地覆盖分类。采用改进网格法和遗传算法对随机森林算法的参数进行选择与优化,比较了改进的网格法和遗传算法方法找到的参数组合最优解,并将优化参数后的随机森林算法与传统的最大似然法及未经参数优化的随机森林算法对比。实验结果验证了随机森林算法在土地覆盖分类上的适用性和稳定性,且该文提出的基于参数优化的随机森林算法能得到更高的分类精度。
According to the fact that in land cover classification there is no suitable method for determining the parameter combination of random forests algorithm to get the optimal classification result,this paper proposed two methods for optimizing the parameters of random forests algorithm.Taking Changping district,Beijing as the study area,this paper employed random forests to perform land cover classification based on spectral,texture and thematic features.Parameters of random forests algorithm were selected and optimized by the improved grid method and genetic algorithm.After comparing the two parameter optimization solutions,this paper compared the land cover classification results by parameters optimization based random forests algorithm with those by traditional maximum likelihood method and random forest without parameter optimization.The experimental results verified the applicability and stability of random forests algorithm on land cover classification and showed that the proposed method has higher classification accuracy.
According to the fact that in land cover classification there is no suitable method for determining the parameter combination of random forests algorithm to get the optimal classification result,this paper proposed two methods for optimizing the parameters of random forests algorithm.Taking Changping district,Beijing as the study area,this paper employed random forests to perform land cover classification based on spectral,texture and thematic features.Parameters of random forests algorithm were selected and optimized by the improved grid method and genetic algorithm.After comparing the two parameter optimization solutions,this paper compared the land cover classification results by parameters optimization based random forests algorithm with those by traditional maximum likelihood method and random forest without parameter optimization.The experimental results verified the applicability and stability of random forests algorithm on land cover classification and showed that the proposed method has higher classification accuracy.
引文
[1]CHAN J C W,PAELINCKX D.Evaluation of random forest and adaboost tree-based ensemble classification and spectral band selection for ecotope mapping using airborne hyperspectral imagery[J].Remote Sensing of Environment,2008,112(6):2999-3011.
[2]GHIMIRE B,ROGAN J,MILLER J.Contextual landcover classification:incorporating spatial dependence in land-cover classification models using random forests and the getis statistic[J].Remote Sensing Letters,2010,1(1):45-54.
[3]LAWRENCE R L,WOOD S D,SHELEY R L.Mapping invasive plants using hyperspectral imagery and breiman cutler classifications(randomforest)[J].Remote Sensing of Environment,2006,100(3):356-362.
[4]RODRIGUEZ-GALIANO V F,GHIMIRE B,ROGAN J,et al.An assessment of the effectiveness of a random forest classifier for land-cover classification[J].ISPRS Journal of Photogrammetry and Remote Sensing,2012,67(1):93-104.
[5]BREIMAN L.Random forests[J].Machine Learning,2001,45(1):5-32.
[6]李治,杨晓梅,孟樊,等.物候特征辅助下的随机森林宏观尺度土地覆盖分类方法研究[J].遥感信息,2013,28(6):48-55.(LI Zhi,YANG Xiaomei,MENG Fan,et al.LULC classification based on random forest with the aid of phenological features[J].Remote Sensing Information,2013,28(6):48-55.)
[7]GISLASON P O,BENEDIKTSSON J A,SVEINSSON J R.Random forests for land cover classification[J].Pattern Recognition Letters,2006,27(4):294-300.
[8]BREIMAN L,FRIEDMAN J,STONE C J,et al.Classification and regression trees[M].BocaRaton,Florida:CRC Press,1984:103-105.
[9]BREIMAN L.Bagging predictors[J].Machine Learning,1996,24(2):123-140.
[10]FELLER W.An introduction to probability theory and its applications[M].New York:John Wiley&Sons,2008:120-123.
[11]申慧珺,席慧,谢刚.改进的网格搜索算法在SVM故障诊断中的应用[J].机械工程与自动化,2012(2):108-110.(SHEN Huijun,XI Hui,XIE Gang.Application of improved grid search algorithm on SVM for fault diagnosis[J].Mechanical Engineering&Automation,2012(2):108-110.)
[12]术洪磊,毛赞猷.GIS辅助下的基于知识的遥感影像分类方法研究:以土地覆盖/土地利用类型为例[J].测绘学报,1997,26(4):328-336.(SHU Honglei,MAO Zanyou.Knowledge based image classification approach supported by a GIS[J].Acta Geodaetica et Cartographica Sinica,1997,26(4):328-336.)
[13]金飞,王番,刘智,等.基于对数极坐标变换的遥感影像居民地提取[J].测绘科学,2013,38(3):102-105.(JIN Fei,WANG Fan,LIU Zhi,et al.Extraction of residential areas from remote sensing images based on log-polar transformation[J].Science of Surveying and Mapping,2013,38(3):102-105.)
[14]冯建辉,杨玉静.基于灰度共生矩阵提取纹理特征图像的研究[J].北京测绘,2007(3):19-22.(FENG Jianhui,YANG Yujing.Study of texture images extraction based on gray level co-occurrence matrix[J].Beijing Surveying and Mapping,2007(3):19-22.)
[15]MCFEETERS S K.The use of the normalized difference water index(NDWI)in the delineation of open water features[J].International Journal of Remote Sensing,1996,17(7):1425-1432.
[16]徐涵秋.利用改进的归一化差异水体指数(MNDWI)提取水体信息的研究[J].遥感学报,2005,9(5):589-595.(XU Hanqiu.A study on information extraction of water body with the modified normalized difference water index(MNDWI)[J].Journal of Remote Sensing,2005,9(5):589-595.)
[17]赵英时.美国中西部沙山地区环境变化的遥感研究[J].地理研究,2001,20(2):213-219.(ZHAO Yingshi.A study on environmental change analysis in sand hill of nebraska using remote sensing[J].Geographical Research,2001,20(2):213-219.)
[18]查勇,倪绍祥,杨山.一种利用TM图像自动提取城镇用地信息的有效方法[J].遥感学报,2003,7(1):37-40.(ZHA Yong,NI Shaoxiang,YANG Shan.An effective approach to automatically extract urban land-use from TM imagery[J].Journal of Remote Sensing,2003,7(1):37-40.)
[19]PAL M,MATHER P M.An assessment of the effectiveness of decision tree methods for land cover classification[J].Remote Sensing of Environment,2003,86(4):554-565.
[20]PAUL E U,CARLA E B.An incremental method for finding multivariate splits for decision trees[C]//Proceedings of the Seventh International Conference on Machine Learning,USA.[S.l.]:[s.n.],1990:58-65.
[21]曹江丽,孙潮义.多目标遗传算法在水下机器人路径规划中的应用[J].武汉大学学报:信息科学版,2010,35(4):441-445.(CAO Jiangli,SUN Chaoyi.Path planning based on multi-objective genetic algorithm for AUV on VCF electronic chart[J].Geomatics and Information Science of Wuhan University,2010,35(4):441-445.)
[2]GHIMIRE B,ROGAN J,MILLER J.Contextual landcover classification:incorporating spatial dependence in land-cover classification models using random forests and the getis statistic[J].Remote Sensing Letters,2010,1(1):45-54.
[3]LAWRENCE R L,WOOD S D,SHELEY R L.Mapping invasive plants using hyperspectral imagery and breiman cutler classifications(randomforest)[J].Remote Sensing of Environment,2006,100(3):356-362.
[4]RODRIGUEZ-GALIANO V F,GHIMIRE B,ROGAN J,et al.An assessment of the effectiveness of a random forest classifier for land-cover classification[J].ISPRS Journal of Photogrammetry and Remote Sensing,2012,67(1):93-104.
[5]BREIMAN L.Random forests[J].Machine Learning,2001,45(1):5-32.
[6]李治,杨晓梅,孟樊,等.物候特征辅助下的随机森林宏观尺度土地覆盖分类方法研究[J].遥感信息,2013,28(6):48-55.(LI Zhi,YANG Xiaomei,MENG Fan,et al.LULC classification based on random forest with the aid of phenological features[J].Remote Sensing Information,2013,28(6):48-55.)
[7]GISLASON P O,BENEDIKTSSON J A,SVEINSSON J R.Random forests for land cover classification[J].Pattern Recognition Letters,2006,27(4):294-300.
[8]BREIMAN L,FRIEDMAN J,STONE C J,et al.Classification and regression trees[M].BocaRaton,Florida:CRC Press,1984:103-105.
[9]BREIMAN L.Bagging predictors[J].Machine Learning,1996,24(2):123-140.
[10]FELLER W.An introduction to probability theory and its applications[M].New York:John Wiley&Sons,2008:120-123.
[11]申慧珺,席慧,谢刚.改进的网格搜索算法在SVM故障诊断中的应用[J].机械工程与自动化,2012(2):108-110.(SHEN Huijun,XI Hui,XIE Gang.Application of improved grid search algorithm on SVM for fault diagnosis[J].Mechanical Engineering&Automation,2012(2):108-110.)
[12]术洪磊,毛赞猷.GIS辅助下的基于知识的遥感影像分类方法研究:以土地覆盖/土地利用类型为例[J].测绘学报,1997,26(4):328-336.(SHU Honglei,MAO Zanyou.Knowledge based image classification approach supported by a GIS[J].Acta Geodaetica et Cartographica Sinica,1997,26(4):328-336.)
[13]金飞,王番,刘智,等.基于对数极坐标变换的遥感影像居民地提取[J].测绘科学,2013,38(3):102-105.(JIN Fei,WANG Fan,LIU Zhi,et al.Extraction of residential areas from remote sensing images based on log-polar transformation[J].Science of Surveying and Mapping,2013,38(3):102-105.)
[14]冯建辉,杨玉静.基于灰度共生矩阵提取纹理特征图像的研究[J].北京测绘,2007(3):19-22.(FENG Jianhui,YANG Yujing.Study of texture images extraction based on gray level co-occurrence matrix[J].Beijing Surveying and Mapping,2007(3):19-22.)
[15]MCFEETERS S K.The use of the normalized difference water index(NDWI)in the delineation of open water features[J].International Journal of Remote Sensing,1996,17(7):1425-1432.
[16]徐涵秋.利用改进的归一化差异水体指数(MNDWI)提取水体信息的研究[J].遥感学报,2005,9(5):589-595.(XU Hanqiu.A study on information extraction of water body with the modified normalized difference water index(MNDWI)[J].Journal of Remote Sensing,2005,9(5):589-595.)
[17]赵英时.美国中西部沙山地区环境变化的遥感研究[J].地理研究,2001,20(2):213-219.(ZHAO Yingshi.A study on environmental change analysis in sand hill of nebraska using remote sensing[J].Geographical Research,2001,20(2):213-219.)
[18]查勇,倪绍祥,杨山.一种利用TM图像自动提取城镇用地信息的有效方法[J].遥感学报,2003,7(1):37-40.(ZHA Yong,NI Shaoxiang,YANG Shan.An effective approach to automatically extract urban land-use from TM imagery[J].Journal of Remote Sensing,2003,7(1):37-40.)
[19]PAL M,MATHER P M.An assessment of the effectiveness of decision tree methods for land cover classification[J].Remote Sensing of Environment,2003,86(4):554-565.
[20]PAUL E U,CARLA E B.An incremental method for finding multivariate splits for decision trees[C]//Proceedings of the Seventh International Conference on Machine Learning,USA.[S.l.]:[s.n.],1990:58-65.
[21]曹江丽,孙潮义.多目标遗传算法在水下机器人路径规划中的应用[J].武汉大学学报:信息科学版,2010,35(4):441-445.(CAO Jiangli,SUN Chaoyi.Path planning based on multi-objective genetic algorithm for AUV on VCF electronic chart[J].Geomatics and Information Science of Wuhan University,2010,35(4):441-445.)