基于对象BOW特征的高分辨率遥感影像变化检测方法

doi:10.12082/dqxxkx.2018.170537

地球信息科学学报 ›› 2018, Vol. 20 ›› Issue (8): 1150-1159.doi: 10.12082/dqxxkx.2018.170537

基于对象BOW特征的高分辨率遥感影像变化检测方法

罗星^1,^2,³(), 徐伟铭^1,^2,^3,^*(), 王佳^1,^2,³

1. 福州大学地理空间信息技术国家地方联合工程研究中心,福州 350116
2. 空间数据挖掘与信息共享教育部重点实验室, 福州 350116
3. 福建省空间信息工程研究中心,福州 350116

收稿日期:2017-11-17 修回日期:2018-04-16 出版日期:2018-08-25 发布日期:2018-08-24
通讯作者: 徐伟铭 E-mail:Luo_Xing_Xing@163.com;xwming2@126.com
作者简介:
作者简介：罗星（1992-）,女,安徽合肥人,硕士生,研究方向为高分辨率遥感影像处理,数据挖掘。E-mail: Luo_Xing_Xing@163.com
基金资助:
福建省科技厅引导性项目（2017Y0055）;海西政务大数据协同创新中心项目;“数字福建”重大项目（闽发改网数字函[2016]203号）

Change Detection Method for High Resolution Remote Sensing Images Based on BOW Features Representation

LUO Xing^1,^2,³(), XU Weiming^1,^2,^3,^*(), WANG Jia^1,^2,³

1. National Engineering Research Centre of Geospatial Information Technology, Fuzhou University, Fuzhou 350116, China
2. Key Lab of Spatial Data Mining & Information Sharing of Ministry of Education, Fuzhou 350116, China;
3. Spatial Information Engineering Research Centre of Fujian Province, Fuzhou 350116, China

Received:2017-11-17 Revised:2018-04-16 Online:2018-08-25 Published:2018-08-24
Contact: XU Weiming E-mail:Luo_Xing_Xing@163.com;xwming2@126.com
Supported by:
Science and Technology Agency of Fujian Province, No.2017Y0055;Haixi Government Affairs Big Data Collaborative Innovation Center;“Digital Fujian” Program, No.2016-203.

摘要/Abstract

摘要：

针对传统基于像素的变化检测方法的缺点,以及底层特征表现能力不足等问题,提出一种基于对象BOW特征的变化检测方法。首先,将经过预处理操作的两期影像进行波段组合得到组合后影像,再考虑地物光谱特征和几何空间信息对组合后影像进行多尺度分割,获得相对应的对象基元;同时,分别提取两幅影像的底层特征（包括影像各波段的均值和方差以及灰度图像的6种纹理特征）。其次,将对象视作文档,像素的特征向量视作单词,利用BOW模型构建影像对象的中层表达,即对象的BOW特征。最后,通过相似性度量算法比较相应对象的BOW特征,从而识别出影像上的变化区域。本文利用2组WorldView-2影像进行了检验,结果表明本文方法的变化检测结果较为完整,精度优于对比方法。本文方法基本能够满足变化检测的需求,为高分辨率遥感影像上的数据挖掘分析提供了有效的手段。

关键词: 视觉词包模型, 变化检测, 中层特征, 对象, 高分辨率遥感影像

Abstract:

A novel change detection method is proposed based on BOW (Bag of Words) features of the image objects, aiming at the drawbacks of traditional change detection methods and the poor interpretation capacity of basic features. First of all, a simultaneous segmentation method which took both spectral features and geometric information into consideration was applied to two images that were previously preprocessed and layer-stacked to acquire corresponding image objects; then mathematical mean and variance of every image band, and six textures from its panchromatic image were extracted as basic features. Moreover, the BOW representation of each image object was constructed by regarding objects as documents and basic features of pixels as words. Finally, similarity measurement was used to compare BOW features to define changed and no-changed areas. In this paper, experimental results on two sets of WorldView-2 images showed that the proposed method demonstrated superiority in completeness of detection result and outperformed the methods for comparison in accuracy assessment. In all, the proposed method can basically meet the needs of change detection work and provide an effective way for data mining and analysis of high resolution remote sensing images.

Key words: bag-of-visual-words, change detection, mid-features, objects, high resolution remote sensing images

罗星, 徐伟铭, 王佳. 基于对象BOW特征的高分辨率遥感影像变化检测方法[J]. 地球信息科学学报, 2018, 20(8): 1150-1159.DOI:10.12082/dqxxkx.2018.170537

LUO Xing,XU Weiming,WANG Jia. Change Detection Method for High Resolution Remote Sensing Images Based on BOW Features Representation[J]. Journal of Geo-information Science, 2018, 20(8): 1150-1159.DOI:10.12082/dqxxkx.2018.170537

图/表 11

图1

图2

图3

图4

图5

图6

图7

图8

表1

图 9

表2

参考文献 15

[16]	赵理君,唐娉,霍连志,等.图像场景分类中视觉词包模型方法综述[J].中国图象图形学报,2014,19(3):333-342. doi: 10.11834/jig.20140301
	[ Zhao L J, Tang P, Huo L Z, et al.Review of the bag-of-visual-words models in image scene classification[J]. Journal of Image and Graphics, 2014,19(3):333-342. ] doi: 10.11834/jig.20140301
[17]	Leung T, Malik J.Representing and recognizing the visual appearance of materials using three-dimensional textons[J]. International Journal of Computer Vision, 2001,43(1):29-44. doi: 10.1023/A:1011126920638
[18]	Huang X, Liu H, Zhang L.Spatiotemporal detection and analysis of urban villages in Mega City Regions of China using high-resolution remotely sensed imagery[J]. IEEE Transactions on Geoscience & Remote Sensing, 2015,53(7):3639-3657.
[19]	Wen D, Huang X, Zhang L, et al.A novel automatic change detection method for urban high-resolution remotely sensed imagery based on multiindex scene representation[J]. IEEE Transactions on Geoscience & Remote Sensing, 2015,54(1):609-625.
[20]	Wu C, Zhang L F, Zhang L P.A scene change detection framework for multi-temporal very high resolution remote sensing images[J]. Signal processing, 2016,124:184-197. doi: 10.1016/j.sigpro.2015.09.020
[21]	Shao W, Yang W, Xia G S, et al.A hierarchical scheme of multiple feature fusion for high-resolution satellite scene categorization[C]. Computer Vision Systems, Germany: Springer Berlin Heidelberg, 2013:324-333.
[22]	Haralick R M.Texture features for image classification[J]. IEEE Transactions on Systems Man and Cybernetics, 1973,3(6):610-621. doi: 10.1109/TSMC.1973.4309314
[1]	Lu D, Mausel P, Brondizio E, et al.Change detection techniques[J]. International Journal of Remote Sensing, 2004,25(12):2365-2401. doi: 10.1080/0143116031000139863
[2]	林祥国,张继贤.面向对象的形态学建筑物指数及其高分辨率遥感影像建筑物提取应用[J].测绘学报,2017,46(6):724-733.
[23]	徐盛. 基于主题模型的高空间分辨率遥感影像分类研究[D].上海:上海交通大学,2012.
	[ Xu S.Topic model on image classification and their applications in high spatial resolution remote sensing image[D]. Shanghai: Shanghai Jiao Tong University, 2012. ]
[2]	[ Lin X G, Zhang J X.Object-based morphological building index for building extraction from high resolution remote sensing imagery[J]. Acta Geodaetica et Cartographica Sinica, 2017,46(6):724-733. ]
[3]	朱长青,王耀革,马秋禾,等.基于形态分割的高分辨率遥感影像道路提取[J].测绘学报,2004,33(4):347-351.
[24]	孙开敏,陈艳.对象级变化检测中的变化向量分析法[C].The International Conference on Computational Intelligence and Industrial Application, 2010.
	[ Sun K M, Chen Y.The application of objects change vector analysis in object-level change detection[C]. The International Conference on Computational Intelligence and Industrial Application, 2010. ]
[3]	[ Zhu C Q, Wang Y G, Ma Q H, et al.Road extraction from high-resolution remotely sensed image based on morphological segmentation[J]. Acta Geodaetica et Cartographica Sinica, 2004,33(4):347-351. ]
[4]	Hussain M., Chen D, Cheng A, et al.Change detection from remotely sensed images: From pixel-based to object-based approaches[J]. ISPRS Journal of Photogrammetry & Remote Sensing, 2013,80(2):91-106. doi: 10.1016/j.isprsjprs.2013.03.006
[5]	Huang X, Zhang L P, Li P.Classification and extraction of spatial features in urban areas using high-resolution multi-spectral imagery[J]. IEEE Geoscience & Remote Sensing Letters, 2007,4(2):260-264.
[7]	Blaschke T.Object based image analysis for remote sensing[J]. ISPRS Journal of Photogrammetry & Remote Sensing, 2010,65(1):2-16. doi: 10.1016/j.isprsjprs.2009.06.004
[8]	刘纯,洪亮,陈杰,等.融合像素—多尺度区域特征的高分辨率遥感影像分类算法[J].遥感学报,2015,19(2):228-239. doi: 10.11834/jrs.20154035
	[ Liu C, Hong L, Chen J, et al.Fusion of pixel-based and multi-scale region-based features for the classification of high-resolution remote sensing image[J]. Journal of Remote Sensing, 2015,19(2):228-239. ] doi: 10.11834/jrs.20154035
[9]	崔林丽,唐娉,赵忠明.一种基于对象和多种特征整合的分类识别方法研究[J].遥感学报,2006,10(1):104-110. doi: 10.3321/j.issn:1007-4619.2006.01.016
	[ Cui L L, Tang P, Zhao Z M.Study on object-oriented classification method by integrating various features[J]. Journal of Remote Sensing, 2006,10(1):104-110. ] doi: 10.3321/j.issn:1007-4619.2006.01.016
[10]	Hazel G G.Object-level change detection in spectral imagery[J]. IEEE Transactions on Geoscience & Remote Sensing, 2001,39(3):553-561. doi: 10.1109/36.911113
[11]	佃袁勇,方圣辉,姚崇怀.一种面向地理对象的遥感影像变化检测方法[J].武汉大学学报·信息科学版,2014,39(8):906-912. doi: 10.13203/j.whugis20130053
	[ Dian Y Y, Fang S H, Yao C H.The geographic object-based method for change detection with remote sensing imagery[J]. Geomatics and Information Science of Wuhan University, 2014,39(8):906-912. ] doi: 10.13203/j.whugis20130053
[12]	王琰,舒宁,龚龑.高分辨率遥感影像土地利用变化检测方法研究[J].国土资源遥感,2012,24(1):43-47. doi: 10.6046/gtzyyg.2012.01.08
	[ Wang Y, Shu N, Gong Y.A study of land use change detection based on high resolution remote sensing images[J]. Remote Sensing of Land and Resources, 2012,24(1):43-47. ] doi: 10.6046/gtzyyg.2012.01.08
[13]	冯文卿,张永军.利用多尺度融合进行面向对象的遥感影像变化检测[J].测绘学报,2015,44(10):1142-1151. doi: 10.11947/j.AGCS.2015.20140260
	[ Feng W Q, Zhang Y J.Object-oriented change detection for remote sensing images based on multi-scale fusion[J]. Acta Geodaetica et Cartographica Sinica, 2015,44(10):1142-1151. ] doi: 10.11947/j.AGCS.2015.20140260
[14]	方涛,霍宏,马贺平.高分辨率遥感影像智能解译[M].北京:科学出版社.2016.
	[ Fang T, Huo H, Ma H P.Intelligent interpretation of high resolution remote sensing images[M]. Beijing: Science Press, 2016. ]
[15]	闫利,朱睿希,刘异,等.顾及遥感影像场景类别信息的视觉单词优化分类[J].遥感学报,2017,21(2):280-290.
	[ Yan L, Zhu R X, Liu Y, et al.Scene classification of remote sensing images by optimizing visual vocabulary concerning scene label information[J]. Remote Sensing of Land and Resources, 2017,21(2):280-290. ]

实验区	字典大小/k		检测改变	检测未变	正确率/%	虚警率/%	漏检率/%
实验1	50	实际改变	43 601	13 524	91.06	6.43	23.67
	50	实际未变	21 558	313 726	91.06	6.43	23.67
	200	实际改变	43 550	13 575	91.82	5.52	23.76
	200	实际未变	185 19	316 765	91.82	5.52	23.76
	300	实际改变	45 092	12 033	91.58	6.26	21.06
	300	实际未变	20 995	314 289	91.58	6.26	21.06
	400	实际改变	44 044	13 081	91.87	5.61	22.90
	400	实际未变	18 824	316 460	91.87	5.61	22.90
	500	实际改变	44 283	12 842	91.41	6.22	22.48
	500	实际未变	20 850	314 434	91.41	6.22	22.48
	600	实际改变	44 044	13 081	91.46	6.10	22.90
	600	实际未变	20 440	314 844	91.46	6.10	22.90
	700	实际改变	43 561	13 564	91.63	5.75	23.74
	700	实际未变	19 289	315 995	91.63	5.75	23.74
	800	实际改变	44 283	12 842	91.41	6.22	22.48
	800	实际未变	20 850	314 434	91.41	6.22	22.48
	1000	实际改变	43 366	13 759	91.10	6.31	24.09
	1000	实际未变	21 147	314 137	91.10	6.31	24.09
实验2	50	实际改变	13 630	6397	84.95	10.65	31.94
	50	实际未变	8177	68 628	84.95	10.65	31.94
	200	实际改变	12 698	7329	85.89	8.24	36.60
	200	实际未变	6331	70 474	85.89	8.24	36.60
	300	实际改变	12 251	7776	85.79	7.79	38.83
	300	实际未变	5984	70 821	85.79	7.79	38.83
	400	实际改变	12 401	7626	85.70	8.10	38.08
	400	实际未变	6221	70 584	85.70	8.10	38.08
	500	实际改变	11 442	8585	85.22	7.45	42.87
	500	实际未变	5723	71 082	85.22	7.45	42.87
	600	实际改变	12 685	7342	85.62	8.57	36.67
	600	实际未变	6580	70 225	85.62	8.57	36.67
	700	实际改变	12 600	7427	85.84	8.19	37.08
	700	实际未变	6288	70 517	85.84	8.19	37.08
	800	实际改变	11 257	8770	85.67	6.65	43.79
	800	实际未变	5108	71 697	85.67	6.65	43.79
	1000	实际改变	12 597	7430	85.72	8.33	37.10
	1000	实际未变	6395	70 410	85.72	8.33	37.10

实验区	方法	正确率/%	虚警率/%	漏检率/%
实验1数据	基于底层特征	68.16	21.65	90.00
实验1数据	OCVA	86.61	7.82	46.08
实验2数据	基于底层特征	82.74	6.83	57.25
实验2数据	OCVA	83.49	8.32	47.92

[1]	韦原原, 江南, 陈云海, 李响, 杨振凯. 顾及时空对象空间相互作用的疫情风险评估建模与应用[J]. 地球信息科学学报, 2021, 23(2): 274-283.
[2]	华一新, 张江水, 曹一冰. 基于时空域的全空间数字世界时空对象组织与管理研究[J]. 地球信息科学学报, 2021, 23(1): 76-83.
[3]	曾梦熊, 华一新, 张江水, 曹一冰, 张政. 多粒度时空对象动态行为表达模型与方法研究[J]. 地球信息科学学报, 2021, 23(1): 104-112.
[4]	李锐, 石佳豪, 董广胜, 刘朝辉. 多粒度时空对象组成结构表达研究[J]. 地球信息科学学报, 2021, 23(1): 113-123.
[5]	张正方, 闫振军, 王增杰, 傅蓉, 罗文, 俞肇元. 基于Bayes网络的多粒度时空对象地理过程演化建模——以新安江模型为例[J]. 地球信息科学学报, 2021, 23(1): 124-133.
[6]	谢雨芮, 江南, 赵文双, 郝睿. 基于多粒度时空对象的作战实体对象化建模研究[J]. 地球信息科学学报, 2021, 23(1): 84-92.
[7]	陈昂, 杨秀春, 徐斌, 金云翔, 张文博, 郭剑, 邢晓语, 杨东. 基于面向对象与深度学习的榆树疏林识别方法研究[J]. 地球信息科学学报, 2020, 22(9): 1897-1909.
[8]	杜培军, 王欣, 蒙亚平, 林聪, 张鹏, 卢刚. 面向地理国情监测的变化检测与地表覆盖信息更新方法[J]. 地球信息科学学报, 2020, 22(4): 857-866.
[9]	张志慧, 刘雯, 李笑含, 朱靖轩, 张洪涛, 杨东, 徐超昊, 徐宪立. 基于无人机航摄影像的喀斯特地区裸岩信息提取及景观格局分析[J]. 地球信息科学学报, 2020, 22(12): 2436-2444.
[10]	吴瑞娟, 何秀凤, 王静. 结合像元级与对象级的滨海湿地变化检测方法[J]. 地球信息科学学报, 2020, 22(10): 2078-2087.
[11]	叶杰, 孟凡晓, 白潍铭, 张斌, 郑金明. “四同”条件下周口城区高分一号遥感影像分类对比研究[J]. 地球信息科学学报, 2020, 22(10): 2088-2097.
[12]	蔡博文,王树根,王磊,邵振峰. 基于深度学习模型的城市高分辨率遥感影像不透水面提取[J]. 地球信息科学学报, 2019, 21(9): 1420-1429.
[13]	张寅丹,王苗苗,陆海霞,刘勇. 基于监督与非监督分割评价方法提取高分辨率遥感影像特定目标地物的对比研究[J]. 地球信息科学学报, 2019, 21(9): 1430-1443.
[14]	李泽宇,明冬萍,范莹琳,赵林峰,刘思民. 遥感影像监督分割评价指标比较与分析[J]. 地球信息科学学报, 2019, 21(8): 1265-1274.
[15]	张俊瑶,姚永慧,索南东主,郜丽静,王晶,张兴航. 基于垂直带谱的太白山区山地植被遥感信息提取[J]. 地球信息科学学报, 2019, 21(8): 1284-1294.

基于对象BOW特征的高分辨率遥感影像变化检测方法

Change Detection Method for High Resolution Remote Sensing Images Based on BOW Features Representation

RichHTML

PDF (PC)

赞

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 15

相关文章 15

Metrics

本文评价

推荐阅读 0