主题模型与SVM组合的小尺度街区用地分类方法

doi:10.12082/dqxxkx.2018.170233

摘要/Abstract

摘要：

城市用地分类作为城市规划重要的基础和依据,对城市资源配置、建设发展等具有重要意义。现有研究对用地分类都聚焦于“稀路网、大街区”的大尺度区域,随着城市发展,大尺度区域的规划模式造成了城市交通效率低下、土地低效率开发等问题,而小尺度的城市街区建设规划为解决上述问题提供了一种新的思路。为了充分挖掘现有交通大数据的潜在价值,更以高效地服务小尺度街区规划,本文将主题模型与支持向量机（SVM）相组合,提出一种面向小尺度街区的用地分类方法。本文以上海市黄浦区人民广场附近为研究区域,依据精细路网对研究区域划分,通过对一周（7天）出租车GPS数据处理,结合区域兴趣点（POI）数据,基于隐含狄利克雷分布（LDA）模型和SVM模型进行用地分类。在人工解译的分类图的基础上对本文方法进行精度评价,并基于百度地图地理数据进行结果验证。研究表明本文方法基于现有的交通大数据,实现了对小尺度街区用地分类,方法分类精度较高,在一定程度上可以节约人力物力,以便更好地服务于小尺度城市规划。

关键词: 主题模型, 用地分类, 出租车, 小尺度, 兴趣点, SVM

Abstract:

Urban land classification is the foundation of urban planning, whose result is of great significance to the allocation of urban resources and the development of urban construction. Previous researches of urban land classification are mainly focused on the study of macro-scale areas, which is characterized by “sparse road network and large block system”. However, with the development of cities, the planning model featured by macro-scale area has caused problems such as the low efficiency of urban traffic and land development. To solve these problems, the construction of urban blocks with small scales was put forward. To make full use of the potential value of the current big data of traffic in the block planning with small scales, this paper presents a land classification method for blocks with small scales through combining the topic modeling and support vector machine (SVM). The regions near People's Square of Huangpu District in Shanghai was taken as the study area. We firstly divided the study area according to fine road network, and then formed a regional mobility pattern through processing the data on the GPS of taxis in one week. By using the data on points of interest (POI), the model of Latent Dirichlet Allocation (LDA) and SVM model, the land use classes are identified. Accuracy assessment of the proposed method has been made based on classification map visually interpreted, and the obtained result has been approved by the geographic data of Baidu Map. The results indicated that this method enables the possibility of the land classification of small-scaled blocks, and could achieve high classification accuracy by utilizing the big data of traffic.

Key words: topic model, land use classification, float car, small scale, POI, SVM

文聪聪, 彭玲, 杨丽娜, 池天河. 主题模型与SVM组合的小尺度街区用地分类方法[J]. 地球信息科学学报, 2018, 20(2): 167-175.DOI:10.12082/dqxxkx.2018.170233

WEN Congcong,PENG Ling,YANG Lina,CHI Tianhe. Topic Model Combined with the SVM for Small Scale Land Use Classification[J]. Journal of Geo-information Science, 2018, 20(2): 167-175.DOI:10.12082/dqxxkx.2018.170233

图/表 15

图1

表1

图2

图3

图4

图5

表2

图6

图7

表3

图8

图9

图10

图11

图12

参考文献 26

[1]	周杨杨. 基于城乡统筹规划的土地分类统一性的研究[D].广州:华南理工大学,2010.
	[ Zhou Y Y.Study on the unity of land use classification based on urban and rural planning[D]. Guangzhou: South China University of Technology, 2010. ]
[2]	张丽琴. 城市地区土地利用变化人类活动驱动机制研究——以武汉地区为例[D].武汉:中国地质大学(武汉),2005.
	[ Zhang L Q.Study on driving mechanism of human behavior for urban region land use change: Take Wuhan region as a case[D].Wuhan: China University of Geosciences (Wuhan), 2005. ]
[3]	Yuan J, Zheng Y, Xie X.Discovering regions of different functions in a city using human mobility and POIs[C]. The 18th ACM SigKdd Conference on Knowledge Discovery and Data Mining, Beijing, China: August 12-16, 2012.
[4]	Pan G, Qi G, Zhang D, et al.Land-Use classification using taxi GPS traces[J]. IEEE Transactions on Intelligent Transportation Systems, 2013,14(1):113-123. doi: 10.1109/TITS.2012.2209201
[5]	梁军辉,林坚,杜洋.大数据条件下城市用地类型辨识研究——基于出租车GPS数据的动态感知[J].上海国土资源,2016,37(1):28-32.
	[ Liang J H, Lin J, Du Y.Research on the identification of urban land use under the big data condition: dynamic awareness based on taxi GPS data[J]. Shanghai Land & Resources, 2016,37(1):28-32. ]
[6]	Jacobs J.The Death and Life of Great American Cities[M]. New York: Random House, 1961.
[7]	赵燕菁. 从计划到市场:城市微观道路-用地模式的转变[J].城市规划,2002,26(10):24-30. doi: 10.3321/j.issn:1002-1329.2002.10.005
	[ Zhao Y J.From planning to marketing: The shift of the road-land use mode of urban micro-roads[J]. Urban Planning, 2002,26(10):24-30. ] doi: 10.3321/j.issn:1002-1329.2002.10.005
[8]	王轩轩,许险峰.可持续发展的小街区模式优势与规划设计原则探讨[C].中国城市规划年会,2009.
	[ Wang X X, Xu X F.Discussion on the sustainable development pattern of small blocks and the principles of planning and design[C]. Annual meeting of urban planning in China, 2009. ]
[9]	Akanwa A O, Okeke F I, Nnodu V C, et al.Quarrying and its effect on vegetation cover for a sustainable development using high-resolution satellite image and GIS[J]. Environmental Earth Sciences, 2017,76(14):505. doi: 10.1007/s12665-017-6844-x
[10]	Gurgel R S, Farias P R S, Oliveira S N D. Land use and land cover mapping and identification of misuse in the permanent preservation areas in the Tailândia Municipality - PA[J]. Semina Ciencias Agrarias, 2017,38(3):1145. doi: 10.5433/1679-0359.2017v38n3p1145
[11]	李英利,闫会杰,张宁丽,等.基于高分遥感影像地表覆盖草地信息人工解译技术方法[J].测绘技术装备,2014(1):67-69.
	[ Li Y L, Yan H J, Zhang N L, et al.Artificial interpretation technique of land cover information of the grassland based on high resolution remote sensing images[J]. Geomatics Technology and Equipment, 2014(1):67-69. ]
[12]	李英利,闫会杰,苟卫涛,等.地理国情普查地表覆盖林地信息人工解译技术方法探讨[J].测绘标准化,2014(1): 4-6.
	[ Li Y L, Yan H J, Gou W T, et al.Discussion on artificial interpretation technology of land cover information of forests in national geographic conditions census[J]. Standardization of Surveying and Mapping, 2014(1):4-6. ]
[13]	严枫,刘登忠,汪友明. QuickBird影像目视解译在土地利用类型调查中的应用——以遂宁市船山区新桥镇为例[J].新疆环境保护,2008,30(1):6-10. doi: 10.3969/j.issn.1008-2301.2008.01.002
	[ Yan F, Liu D Z, Wang Y M.Application of visual interpretation of QuickBird Images for land-use types investigation: A case study of Xinqiao Town, Chuanshan District in Suining City[J]. Environmental Protection of Xinjiang, 2008,30(1):6-10. ] doi: 10.3969/j.issn.1008-2301.2008.01.002
[14]	D. Blei.Introduction to probabilistic topic models[C]. Communications of the ACM, 2011.
[15]	彭璐. 支持向量机分类算法研究与应用[D].长沙:湖南大学,2007.
	[ Peng L.Research on classification algorithm of support vector machine and its applications[D]. Changsha: Hunan University, 2007. ]
[16]	Bengio Y, Gr Y.No unbiased estimator of the variance of K-Fold cross-kalidation[J]. Journal of Machine Learning Research, 2003,5(22):1089-1105. doi: 10.1115/1.1789979
[17]	Blei D, Ng A, Jordan M.Latent Dirichlet Allocation[J]. Journal of Machine Learning Research 2003,3:993-1022
[18]	Geman S, Geman D. Stochastic Relaxation, Gibbs Distributions,the Bayesian Restoration of Images[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 1984,6(6):721-41.
[19]
[20]	池娇,焦利民,董婷,等.基于POI数据的城市功能区定量识别及其可视化[J].测绘地理信息, 2016,41(2):68-73.
	[ Chi J,, Jiao L M,, Dong T, et al.Quantitative identification and visualization of urban functional area based on POI data[J].Journal of Geomatics, 2016,41(2):68-73. ]
[21]	赵卫锋,李清泉,李必军.利用城市POIs数据提取分层地标[J].遥感学报,2011,15(5):981-989. doi: 10.11834/jrs.20110173
	[ Zhao W F, Li Q Q, Li B J.Extraction hierarchical landmarks using urban POIs data[J]. Journal Of Remote Sensing, 2011,15(5):981-989. ] doi: 10.11834/jrs.20110173
[22]	Cortes C, Vapnik V.Support-vector networks[J]. Machine Learning, 1995,20:273-297.
[23]	奉国和. SVM分类核函数及参数选择比较[J].计算机工程与应用,2011,47(3):123-124. doi: 10.3778/j.issn.1002-8331.2011.03.037
	[ Feng G H.Parameter optimization for support vector machines classification. Computer Engineering and Applications, 2011,47(3):123-124. ] doi: 10.3778/j.issn.1002-8331.2011.03.037
[24]
[25]	Landis J R, Koch G G.The measurement of observer agreement for categorical data[J]. Biometrics, 1977,33(1):159-174. doi: 10.2307/2529310
[26]	王劲峰,葛咏,李连发,等.地理学时空数据分析方法[J].地理学报,2014,69(9):1326-1345. doi: 10.11821/dlxb201409007
	[ Wang J F, Ge Y, Li L F.Spatiotemporal data analysis in geography[J]. Acta Geographica Sinica, 2014,69(9):1326-1345. ] doi: 10.11821/dlxb201409007

数据	年份	数量	详细说明
上海市区划图	2015		从区划图中选取黄浦区人民广场附近为研究区域
研究区域路网数据/条	2015	508	从上海市全部路网中提取研究区域内的路网,道路类别有一级道路、二级道路、三级道路等13类
研究区域内POI数据/个	2015	4716	从上海市全部POI数据中提取研究区域内的POI,共有交通设施、体育休闲、公司企业等19个类别
研究区域内出租车轨迹数据/万条	2015	734	从上海市全部出租车轨迹数据中提取研究区域内的轨迹数据

方法结果	解译结果
方法结果	居住用地	公共管理与公共服务设施用地	商业服务设施用地	工业用地	道路与交通用地	绿地与广场用地	总计
居住用地	822	10	8	9	20	136	1005
公共管理与公共服务设施用地	1	3	0	0	0	3	7
商业服务设施用地	1	3	62	1	2	17	86
工业用地	1	2	6	1	0	0	10
道路与交通用地	0	0	0	0	8	2	10
绿地与广场用地	2	0	0	4	1	32	39
总计	827	18	76	15	31	190	1157

方法	总体精度	Kappa系数
常规特征非监督分类方法	0.298185	0.125381
LDA模型非监督分类方法	0.413654	0.251723
本文方法	0.802074	0.462550