Winter wheat classification method based on feature optimization of random forest

doi:10.13474/j.cnki.11-2246.2022.0080

Abstract

Abstract: Based on the multi-temporal Landsat 8 OLI data,this paper conducts research on the feature extraction and feature selection methods of comprehensive spectral features and vegetation index features.By analyzing the temporal changes of the spectral and vegetation index features,the optimal time-phase spectrum is extracted,and the wheat extraction features are constructed.A random forest feature selection algorithm based on importance and Pearson correlation is used to select features and classify them.The results show that:when using the selected features to classify,the overall accuracy of classification is 89.78%,and the classification accuracy of wheat is 98.33%.Compared with the classification results of the features before optimization,the classification accuracy is increased by 2.96% and 2.55%,respectively.Random forest feature selection based on importance and relevance not only improves the classification accuracy,but also improves the efficiency of the classifier.

Key words: feature selection;random forest;Pearson correlation;winter wheat extraction

CLC Number:

P237

FENG Zhili, XIAO Feng, LU Xiaoping, HAO Bo, WANG Ruyi, ZHU Rui. Winter wheat classification method based on feature optimization of random forest[J]. Bulletin of Surveying and Mapping, 2022, 0(3): 70-75.

References

[1] 王娜,李强子,杜鑫,等.单变量特征选择的苏北地区主要农作物遥感识别[J].遥感学报, 2017, 21(4):519-530.
[2] 胡琼,吴文斌,宋茜,等.农作物种植结构遥感提取研究进展[J].中国农业科学, 2015, 48(10):1900-1914.
[3] ZHENG Baojuan, MYINT S W, THENKABAIL P S, et al. A support vector machine to identify irrigated crop types using time-series Landsat NDVI data[J]. International Journal of Applied Earth Observation and Geoinformation, 2015, 34:103-112.
[4] FOERSTER S, KADEN K, FOERSTER M, et al. Crop type mapping using spectral-temporal profiles and phenological information[J]. Computers and Electronics in Agriculture, 2012, 89(3):30-40.
[5] 刘吉凯,钟仕全,梁文海.基于多时相Landsat 8 OLI影像的作物种植结构提取[J].遥感技术与应用, 2015, 30(4):775-783.
[6] 马超,杨飞,王学成.基于中尺度光谱和时序物候特征提取南方丘陵山区茶园[J].国土资源遥感, 2019, 31(1):141-148.
[7] 王文静,张霞,赵银娣,等.综合多特征的Landsat 8时序遥感图像棉花分类方法[J].遥感学报, 2017, 21(1):115-124.
[8] 胡琼.基于时序MODIS影像的农作物遥感识别方法研究[D].北京:中国农业科学院, 2018.
[9] 陈安旭,李月臣.基于Sentinel-2影像的西南山区不同生长期水稻识别[J].农业工程学报, 2020, 36(7):192-199.
[10] 刘杰,刘吉凯,安晶晶,等.基于时序Landsat 8 OLI多特征与随机森林算法的作物精细分类研究[J].干旱地区农业研究, 2020, 38(3):281-288.
[11] 王李娟,孔钰如,杨小冬,等.基于特征优选随机森林算法的农耕区土地利用分类[J].农业工程学报, 2020, 36(4):244-250.
[12] 杨蕙宇,王征强,白建军,等.基于多特征提取与优选的冬小麦面积提取[J].陕西师范大学学报(自然科学版), 2020, 48(1):40-49.
[13] 赵晓晴,王萍,荆林海,等.时序Sentinel-2A影像光谱特征的茶园提取应用[J].测绘科学, 2020, 45(6):80-88.
[14] 靳宁,张东彦,李振海,等.基于多光谱卫星模拟波段反射率的冬小麦水分状况评估[J].农业机械学报, 2020, 51(11):243-252.
[15] 张莹.基于多季相Landsat-8 OLI影像的多特征森林植被分类[D].北京:北京林业大学, 2018.
[16] 史蕾.面向对象的高分辨率遥感影像特征选择与分类研究[D].武汉:武汉大学, 2018.
[17] 张磊,宫兆宁,王启为,等. Sentinel-2影像多特征优选的黄河三角洲湿地信息提取[J].遥感学报, 2019, 23(2):313-326.
[18] 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, 67:93-104.
[19] 王猛,张新长,王家耀,等.结合随机森林面向对象的森林资源分类[J].测绘学报, 2020, 49(2):235-244.
[20] FRAIWAN L, LWEESY K, KHASAWNEH N, et al. Automated sleep stage identification system based on time-frequency analysis of a single EEG channel and random forest classifier[J]. Computer Methods and Programs in Biomedicine, 2012, 108(1):10-19.
[21] 侯蒙京,殷建鹏,葛静,等.基于随机森林的高寒湿地地区土地覆盖遥感分类方法[J].农业机械学报, 2020, 51(7):220-227.
[22] 张良均,谭立云,刘名军. Python数据分析与挖掘实战[M]. 2版.北京:机械工业出版社, 2020.
[23] EDELMANN D, MÓRI T F, SZÉKELY G J. On relationships between the Pearson and the distance correlation coefficients[J]. Statistics&Probability Letters, 2021, 169:108960.