为了实现基于遥感影像作物的自动分类，发挥遥感技术宏观、迅速的大范围监测特点，本文在遥感影像非监督分类的基础上，提出了一种基于ISODATA非监督分类结果的自动分类方法。该方法分为ISODATA非监督分类过程和自动分类过程，自动分类过程又可分为冬小麦样本点占比排序和类冬小麦类别确定两个方面。当非监督分类类别设置为40或50类、每类样本数量为4或5类时，冬小麦遥感分类精度较高且分类精度稳定。在200个样本点组合（40个分类类别，每个类别中5个样本点）中，基于ISODATA非监督自动分类结果的总体精度相较于最大似然分类方法提高了2.5个百分点，KAPPA系数提高了19.4%。在500个样本点组合（100个分类类别，每个类别中5个样本点）下，基于ISODATA非监督自动分类结果总体精度和KAPPA系数与最大似然分类方法相近。基于ISODATA非监督分类结果的自动分类方法可以在样本量较少时保持较高的分类精度，人机交互少，分类效率高，适用于业务化应用。
Abstract
To achieve the crop automatic classification based on remote sensing images and take advantage of the characteristics of macro, rapid and large scale monitoring, based on the remote sensing image unsupervised classification, an automatic classification method on the basis of ISODATA unsupervised classification was provided. The method had ISODATA unsupervised classification process and automatic classification process. The automatic classification process included the proportion ranking of winter wheat sample and winter wheat identification. When the classification categories of unsupervised classification were set to be 40 or 50 respectively, and each category had 4 or 5 samples, the winter wheat classification accuracy was relatively high and stable. Among the 200 sample point combinations (40 classified categories, each category with 5 sample points), the overall accuracy of winter wheat classification based on automatic ISODATA unsupervised classification method was 2.5 percent higher than that of the maximum likelihood classification method and its KAPPA coefficient rose by 19.4%. Among 500 sample point combinations (100 classified categories, each
category with 5 sample points), the winter wheat overall accuracy and its KAPPA coefficient based on automatic ISODATA unsupervised classification method were close to that of the maximum likelihood classification method. The automatic ISODATA unsupervised classification method has relatively higher classification accuracies with fewer samples, and it is featured with less human- machine interaction and higher classification efficiency. Therefore, this method is suitable for operational applications.
关键词
ISODATA,GF-1,WFV,冬小麦,面积,遥感,识别
{{custom_keyword}} /
Key words
ISODATA, GF-1, WFV, winter;wheat, area, remote;sensing, recognition
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 吴炳方,蒙继华,李强子,等.全球农情遥感速报系统(Crop Watch)新进展[J].地球科学进展,2010,25:1013-1022.
[2] 任建强,陈仲新,唐华俊,等.基于遥感信息与作物生长模型的区域作物单产模拟[J].农业工程学报,2011,27:257-264.
[3] 唐华俊,吴文斌,杨鹏,等.农作物空间格局遥感监测研究进展[J].中国农业科学,2010,43(14):2879-2888.
[4] Liu J, Liu M, Tian H, et al. Spatial and temporal patterns of China’s cropland during 1990-2000: An analysis based on Landsat TM data[J]. Remote Sensing of Environment, 2005, 98: 442-456.
[5] 刘佳,王利民,杨福刚,等.基于HJ时间序列数据的农作物种植面积估算[J].农业工程学报,2015,31(3):199-206.
[6] 钱永兰,杨邦杰,雷廷武.数据融合及其在农情遥感监测中的应用与展望[J].农业工程学报,2004,20(4):286-290.
[7] 王利民,刘佳,杨福刚,等.基于GF-1卫星遥感的冬小麦面积早期识别[J].农业工程学报,2015,31(11):194-201.
[8] Ali M M, Qaseem M S, Rajamani L, et al. Extracting useful rules through improved decision tree induction using information entropy[J].International Journal of Information Sciences Techniques,2013,3(1):27-41.
[9] Dai Q Y, Zhang C P, Wu H. Research of Decision Tree Classification Algorithm in Data Mining[J].International Journal of Database Theory and Application,2016,9(5):1-8.
[10] Otukei J R, Blaschke T, Woldai T, et al. Land cover change assessment using decision trees, support vector machines and maximum likelihood classification algorithms.[J].International Journal of Applied Earth Observation Geoinformation, 2010,12(1):S27-S31.
[11] Zhong, Y F, Zhang L P, Gong W. Unsupervised remote sensing image classification using anartificial immune network[J].International Journal of Remote Sensing 2011,32(19):5461-5483.
[12] Huanhuan C, Runsheng W. Integrating Contexts into the Classification of High-Resolution Remote Sensing Images Using the Bayesian Networks[J].Computer Engineering and Science,2011,33(1):70- 76.
[13] Ghosh A, Mishra N S, Ghosh S. Fuzzy Clustering Algorithms for Unsupervised Change Detection in Remote Sensing Images[J].Information Sciences,2011,181(4):699-715.
[14] Xin S, Chu H, Qian F. A Supervised Classification Method Based on Conditional Random Fields with Multi-scale Region Connection Calculus Model for SAR Image[J].IEEE Geoscience and Remote Sensing Letters,2011,8(3):497-501.
[15] Manakos I, Schneider T, Ammer U. A comparison between the ISODATA and the eCognition classification methods on basis of field data[J]. International Archives of Photogrammetry and Remote Sensing, 2000,Vol.XXXIII,SupplementB7:133-139
[16] Liu W P, Hung C C, Kuo B C. An adaptive clustering algorithm based on the possibility clustering and ISODATA for multispectral image classification[J].Remote Sensing and Spatial Information Sciences,2008, Vol. XXXVII. Part B7:565-568.
[17] 赵泉华,赵雪梅,李玉.结合HMRF模型的模糊ISODATA高分辨率遥感图像分割[J].信号处理,2016,32(02):157-166.
[18] 齐永菊,裴亮,张宗科,魏显虎,王旭.一种模糊聚类的遥感影像分析方法研究[J].测绘科学,2017,42(07):139-146.
[19] Liu Q J, Zhao Z M, Li Y X, et al. Feature selection based on sensitivity analysis of fuzzy ISODATA[J].Neuro computing,2012,86:29-37.
[20] 郭云开,曾繁. 融合增强型模糊聚类遗传算法与ISODATA算法的遥感影像分类[J].测绘通报,2015,(12):23-26.
[21] 马彩虹,戴芹,刘士彬.一种融合PSO和Isodata的遥感图像分割新方法[J].武汉大学学报(信息科学版),2012,37(1):35-38.
[22] 刘扬,王鹏,杨瑞,等.基于OpenMP的遥感影像并行ISODATA聚类研究[J].计算机工程,2016,42(07):238-243+250.
[23] 李平阳,郭品文,国文哲.基于HJ-1A 卫星数据的衡水地区冬小麦面积遥感估算应用[J].气象与减灾研究,2015,38(2):47-54.
[24] 王清川,寿绍文,许敏,等.廊坊市暴雨洪涝灾害风险评估与区划[J].干旱气象,2010,28(4):475-482.
[25] 刘佳,王利民,杨玲波,等.基于6S模型的GF-1卫星影像大气校正及效果[J]. 农业工程学报, 2015, 31(19):159-168.
[26] 王利民, 刘佳, 杨玲波,等. 基于NDVI加权指数的冬小麦种植面积遥感监测[J]. 农业工程学报, 2016, 32(17):127-135.
[27] 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.
[28] Hay A M. The derivation of global estimation from a confusion matrix[J].International Journal of Remote Sensing,1988,9(8):1395-1398.
[29] Congalton R G. A comparison of sampling schemes used in generating error matrices for assessing the accuracy of maps generated from remotely sensing data[J].Photogrammetric Engineering and Remote Sensing,1988,54(5):593-600.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
PDF(5865 KB)
