基于随机森林算法的城市不透水面信息提取——以长春市为例

doi:10.13474/j.cnki.11-2246.2020.0352

测绘通报 ›› 2020, Vol. 0 ›› Issue (11): 43-49.doi: 10.13474/j.cnki.11-2246.2020.0352

基于随机森林算法的城市不透水面信息提取——以长春市为例

常翔宇, 柯长青

南京大学地理与海洋科学学院, 江苏南京 210023

收稿日期:2019-12-09 修回日期:2020-07-06 发布日期:2020-11-30
通讯作者: 柯长青。E-mail:kecq@nju.edu.cn E-mail:kecq@nju.edu.cn
作者简介:常翔宇(1997-),男,硕士,研究方向为卫星测高数据处理与应用。E-mail:1119398084@qq.com
基金资助:
江苏省自然科学基金（BK20180343）

Urban impervious surface information extraction based on random forest algorithm: taking Changchun as an example

CHANG Xiangyu, KE Changqing

School of Geography and Ocean Science, Nanjing University, Nanjing 210023, China

Received:2019-12-09 Revised:2020-07-06 Published:2020-11-30

摘要/Abstract

摘要： 为了快速、准确地掌握不透水面的空间分布及满足动态变化信息现实需求，本文基于多分类器集成学习的思想，引入随机森林算法，以Landsat8影像为数据源，长春市为实验区，选取光谱特征、纹理测度、空间变换后的独立分量等25个特征变量进行分类研究，根据OOB误差进行重要性分析并试验得出最优的分类模型，实现高精度不透水面信息的提取，最后与传统参数分类法进行比较。结果表明：随机森林算法的总体精度可以达到94%，高出最大似然分类法5.9%，支持向量机算法0.77%，Kappa系数为0.914 3，均方根误差为0.104 3，不透水面的提取精度达95.54%，可以精确地得出所需信息，为城市建设与规划提供有效的专题数据。

关键词: 不透水面, 随机森林, 影像分类, 特征提取

Abstract: In order to quickly and accurately grasp the spatial distribution and dynamic change information of impervious surface, based on the idea of multi-classifier ensemble learning, random forest algorithm is introduced in eliis paper. Landsat 8 image is used as data source and Changchun city as experimental area. 25 feature variables, such as spectral indices, texture measures and independent components after spatial transformation are selected to classify. The importance of variables calculated by out of bag error is analyzed and the optimal classification model is obtained through many experiments. The extraction of high-precision impervious surface is also realized. Finally, random forest algorithm is compared with the traditional parameter classifier. The result indicated that the overall accuracy of random forest algorithm can reach 94%, which is higher than 5.9% of maximum likelihood classification, 0.77% of support vector machine algorithm, 0.914 3 of Kappa coefficient and 0.104 3 of root mean square error. The extraction accuracy of impervious surface is 95.54%, which can not only accurately extract impervious surface but also provide effective thematic information for urban construction and planning.

Key words: impervious surface, random forest, image classification, feature extraction

中图分类号:

P237

常翔宇, 柯长青. 基于随机森林算法的城市不透水面信息提取——以长春市为例[J]. 测绘通报, 2020, 0(11): 43-49.

CHANG Xiangyu, KE Changqing. Urban impervious surface information extraction based on random forest algorithm: taking Changchun as an example[J]. Bulletin of Surveying and Mapping, 2020, 0(11): 43-49.

参考文献

[1] ARNOLD JR C L, GIBBONS C J. Impervious surface cove-rage:the emergence of a key environment indicator[J]. Journal of the American Planning Association, 1996, 62(2):243-258.
[2] OKE T R. Boundary layer climates[M]. London:Rou-tledge, 1987.
[3] YU S Q. Study on annual variation of precipitation and urban effect in Beijing[J]. Progress in Natural Science, 2007, 17(5):632-638.
[4] AZAR D, GRAESSER J, ENGSTROM R, et al. Spatial refinement of census population distribution using remotely sensed estimates of impervious surfaces in Haiti[J]. International Journal of Remote Sensing, 2010, 31(21):5635-5655.
[5] 哈尚辰, 阿里木江·卡斯木. 基于土地集约利用水平的城市热岛效应影响因子分析[J]. 冰川冻土, 2016, 38(1):270-278.
[6] 李涵, 李龙, 张婷,等. 徐州市中心城区不透水面时空异质性分析[J]. 长江流域资源与环境, 2019, 28(3):178-190.
[7] 徐涵秋. 基于谱间特征和归一化指数分析的城市建筑用地信息提[J].地理研究,2005,24(2):311-320.
[8] LU D, HERTRIC S, MORAN E. Land cover classification in a complex urban-rural landscape with QuickBird imagery[J]. Photogrammetric Engineering & Remote Sensing, 2010, 76(10):1159-1168.
[9] 孙志英,赵彦锋,陈杰,等.面向对象分类在城市地表不可透水度提取中的应用[J]. 地理科学,2007,27(6):837-842.
[10] WENG Q H, HU X F. Medium spatial resolution satellite imagery for estimating and mapping urban impervious surfaces using LSMA and ANN[J]. IEEE Transactions on Geoscience and Remote Sensing, 2008, 46(8):2397-2406.
[11] BREIMAN L. Random forests[J]. Machine Learning, 2001, 45(1):5-32.
[12] HEREMAN S, ORSHOVEN J V. Machine learning methods for sub-pixel land-cover classification in the spatially heterogeneous region of Flanders(Belgium):a multi-criteria comparison[J]. International Journal of Remote Sensing, 2015,36(11):2934-2962.
[13] FAWAGREH K, GABER M M, ELVAN E. Random forests:from early developments to recent advancements[J]. Systems Science & Control Engineering an Open Access Journal, 2014,2(1):602-609.
[14] GAUTAM V K, GAURAV P K, MURUGAN P, et al. Assessment of surface water dynamicsin bangalore using WRI, NDWI, MNDWI, supervised classification and K-T transformation[J]. Aquatic Procedia, 2015, 4:739-746.
[15] 杨斌, 高桂胜, 王磊,等. 基于GF-1 WFV和Landsat-8 OLI提取植被信息方法比较研究[J]. 测绘工程, 2018, 27(8):7-12.
[16] ZOU T, YANG W, DAI D, et al. Polarimetric SAR image classification using multiple features combination and extremely randomized clustering forests[J]. EURASIP Journal on Advances in Signal Processing, 2010, 49(1). DOI:10.1155/2010/465612.
[17] 邵振峰,张源,周伟琪,等.基于测绘卫星影像的城市不透水面提取[J].地理空间信息,2016,14(7):1-5.
[18] FROHN R C, CHAUDHARY N. Multi-scale image segmentation and object-oriented processing for land cover classification[J]. Geosciences & Remote Sensing, 2008, 45(4):377-391
[19] ZHU Z, CURTIS E W, JOHN R, et al. Assessment of spectral, polarimetric, temporal, and spatial dimensions for urban and periurban land cover classification using Landsat and SAR data[J]. Remote Sensing of Environment, 2012, 117:72-82.
[20] BEIJMA S V, COMBER A, LAMB A. Random forest classification of salt marsh vegetation habitats using quad-polarimetric airborne SAR, elevation and optical RS data[J]. Remote Sensing of Environment, 2014, 149:118-129.
[21] 徐涵秋,王美雅.地表不透水面信息遥感的主要方法分析[J],遥感学报,2016,20(5):1270-1289.
[22] 赵艺淞,杨昆,王保云,等.随机森林在城市不透水面提取中的应用研究[J],云南师范大学学报(自然科学版),2017,37(3):73-78.

基于随机森林算法的城市不透水面信息提取——以长春市为例

Urban impervious surface information extraction based on random forest algorithm: taking Changchun as an example

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	吕峥, 孙群, 温伯威, 马京振. 一种自身全局最优的道路网Stroke生成方法[J]. 测绘通报, 2022, 0(7): 93-99.
[2]	杨俊, 刘灵辉. 加权全极化SAR图像分类下的功能区土地时空变化特征提取[J]. 测绘通报, 2022, 0(7): 138-142,153.
[3]	郁宗桥, 王育红, 刘文宋, 左雨芳, 冯锋. 顾及GF-3影像纹理特征的洪涝范围提取方法[J]. 测绘通报, 2022, 0(5): 26-31.
[4]	赵传武, 吴子若, 张玉环, 郭伟. 基于多时相遥感的鄂尔多斯市不透水面变化分析[J]. 测绘通报, 2022, 0(5): 56-61.
[5]	窦世卿, 陈治宇, 徐勇, 郑贺刚, 苗林林, 宋莹莹. 基于多特征融合与典型降维方法的高光谱影像分类[J]. 测绘通报, 2022, 0(4): 32-36,50.
[6]	肖斌, 周忠赣, 范军林, 徐看, 刘向铜. Sentinel+Landsat融合影像的城市不透水面积参数获取[J]. 测绘通报, 2022, 0(4): 66-71,95.
[7]	冯志立, 肖锋, 卢小平, 郝波, 王如意, 朱睿. 基于随机森林特征优选的冬小麦分类方法[J]. 测绘通报, 2022, 0(3): 70-75.
[8]	王晓晓, 韩留生, 杨骥, 李勇, 张大富, 孙广伟, 范俊甫. Sentinel-2与Landsat 8数据组合下的多特征冬小麦面积提取[J]. 测绘通报, 2022, 0(3): 111-115.
[9]	崔青林, 汪鸣泉, 黄永健. 融合随机森林模型和6种水体指数的上海市水体信息提取[J]. 测绘通报, 2022, 0(2): 106-109.
[10]	高雨, 胡召玲, 樊茹. 高分辨率影像融合算法对滨海湿地土地利用分类精度的影响[J]. 测绘通报, 2022, 0(1): 116-120.
[11]	丁鸽, 燕立爽, 彭健, 朱传广, 黄雪亭. 基于RANSAC算法的隧道点云横断面提取[J]. 测绘通报, 2021, 0(9): 120-123.
[12]	耿万轩, 周维勋, 金双根. 基于LDCNN特征提取的多核SVM高分辨率遥感影像场景分类[J]. 测绘通报, 2021, 0(8): 14-21.
[13]	淳锦, 张新长, 郭海京, 张建国, 金诗程. 基于多源信息与深度学习特征提取的人口空间抽样方法[J]. 测绘通报, 2021, 0(8): 42-47.
[14]	李军, 桑潇, 张成业, 赵伟, 刘新华, 王宏鹏, 王金阳, 李佳瑶, 杨颖. 资源型城市长时间序列土壤含水量变化分析——以锡林浩特市为例[J]. 测绘通报, 2021, 0(7): 17-22,38.
[15]	陶莉, 胡召玲. 利用Sentinel-2A数据提取长江中下游丘陵地带农作物种植信息[J]. 测绘通报, 2021, 0(7): 39-43.