Please wait a minute...
浙江大学学报(工学版)
浙江大学学报(工学版)  2021, Vol. 55 Issue (10): 1847-1855    DOI: 10.3785/j.issn.1008-973X.2021.10.006
计算机技术     
面向对象轮廓约束GGVF Snake的建筑物边界优化方法
常京新1(),高贤君1,2,*(),杨元维1,李少华1,王萍3,4
1. 长江大学 地球科学学院，湖北 武汉 430100
2. 武汉大学 测绘遥感信息工程国家重点实验室，湖北 武汉 430079
3. 中国科学院空天信息创新研究院，北京 100094
4. 海南省地球观测重点实验室，海南 三亚 572029
Building boundary optimization method based on object-oriented contour constraint GGVF Snake model
Jing-xin CHANG1(),Xian-jun GAO1,2,*(),Yuan-wei YANG1,Shao-hua LI1,Ping WANG3,4
1. School of Geosciences, Yangtze University, Wuhan 430100, China
2. State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, China
3. Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
4. Key Laboratory of Earth Observation of Hainan Province, Sanya 572029, China
 全文: PDF(1548 KB)   HTML
摘要：

分析高分辨率遥感影像中建筑物的特征和常用方法提取建筑物存在边界漏检误检导致的边界不规则等问题，提出面向对象轮廓约束广义梯度矢量流（GGVF）Snake模型的建筑物边界优化方法. 在利用分类法获取建筑物轮廓的初始结果基础上，自动提取每个建筑物轮廓线作为GGVF Snake的初始轮廓线，获取各轮廓外接矩形进行对象裁剪，提取每个建筑物的子图对象. 对每个子图对象进行Canny边缘检测，结合Hough变换提取直线特征，输入到广义梯度矢量流模型的迭代求解中快速最小化能量函数，实现对象级建筑物轮廓的精确优化. 实验结果表明，利用提出的方法能够自动获取初始建筑物的轮廓信息，提高优化的自动化程度；基于对象级的边缘检测与直线特征的加入，有助于GGVF Snake快速拟合，准确地平滑建筑物边缘且准确度高. 相对于其他对比方法，本文方法的轮廓优化总体精度和综合值均有提升，可以作为有效提升分类原理获取的建筑物轮廓的优化后处理手段，提高了建筑物提取的精度.
关键词： 高分辨率遥感影像;  建筑物提取;  GGVF Snake;  Hough变换;  Canny算子;  轮廓优化    
Abstract:

An object-oriented contour constrained generalized gradient vector flow (GGVF) Snake model for building boundary optimization method was proposed by analyzing the features of buildings and the problems of conventional methods that the irregular boundary problem caused by false classification is pervasive. Each building boundary was automatically extracted as the initial boundary of GGVF Snake based on the initial building results obtained by classification. The circumscribed rectangle of each building boundary was obtained for object clipping in order to extract each building sub-image object. Canny edge detection was performed on each sub-image object to acquire each building boundary result that will be combined with Hough transform to extract linear features. Then the linear features were inputted into the generalized gradient vector flow model’s iterative solution to minimize the energy function. Then the precise of the object-level building contour was enhanced. The experimental results show that the proposed method can automatically obtain the initial building boundary information and improve the automation degree of optimization. The object-level contour detection and the addition of line features help GGVF Snake quickly fit and accurately smooth the building boundary to enhance the contour accuracy. The overall accuracy and comprehensive value of the proposed method are improved compared with other contour optimization methods. Thus the method can be used as an effective post-processing optimization method for the classification principle to improve the building extraction precision.
Key words: high-resolution remote sensing image    building extraction    GGVF Snake    Hough transform    Canny operator    contour optimization
收稿日期: 2020-11-20 出版日期: 2021-10-27
CLC:  TP 753  
基金资助: 国家自然科学基金资助项目（41872129）；自然资源部地理国情监测重点实验室开放基金资助项目（2020NGCM07）；武汉大学测绘遥感信息工程国家重点实验室开放基金资助项目（18R04）；海南省地球观测重点实验室开放基金资助项目（2020LDE001）
通讯作者: 高贤君     E-mail: 201500671@yangtzeu.edu.cn;junxgao@yangtzeu.edu.cn
作者简介: 常京新（1995—），男，硕士生，从事高分辨率遥感影像智能解译的研究.orcid.org/0000-0002-1086-0608. E-mail: 201500671@yangtzeu.edu.cn
服务　　
把本文推荐给朋友
加入引用管理器
E-mail Alert
作者相关文章　　
常京新
高贤君
杨元维
李少华
王萍

引用本文:

常京新,高贤君,杨元维,李少华,王萍. 面向对象轮廓约束GGVF Snake的建筑物边界优化方法[J]. 浙江大学学报(工学版), 2021, 55(10): 1847-1855.

Jing-xin CHANG,Xian-jun GAO,Yuan-wei YANG,Shao-hua LI,Ping WANG. Building boundary optimization method based on object-oriented contour constraint GGVF Snake model. Journal of ZheJiang University (Engineering Science), 2021, 55(10): 1847-1855.

链接本文:

https://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2021.10.006    或    https://www.zjujournals.com/eng/CN/Y2021/V55/I10/1847

图 1  提出方法中间结果的示例
图 2  建筑物轮廓优化的流程图
图 3  影像#1的建筑物轮廓优化方法对比
图 4  影像#2的建筑物轮廓优化方法对比
图 5  影像#3的建筑物轮廓优化方法对比
影像编号 对比方法 CM/% CR/% F1/% OA/% CA/%
#1 文献[8]的建筑物提取结果 92.54 98.83 95.58 91.54 15.54
#1 文献[4]的建筑物提取结果 94.82 95.67 95.24 90.93 19.10
#1 传统GGVF Snake优化文献[4]的结果 94.11 97.69 95.87 92.07 24.30
#1 文献[5]优化文献[4]的结果 94.63 96.82 95.71 91.78 23.86
#1 本文方法优化文献[4]的结果 95.81 98.88 97.32 94.78 29.81
#1 U-Net神经网络提取结果 94.85 96.26 95.55 91.49 16.16
#1 本文方法优化U-Net结果 99.36 96.06 97.68 95.47 41.49
#2 文献[8]的建筑物提取结果 89.64 90.73 90.18 82.11 23.73
#2 文献[4]的建筑物提取结果 85.44 93.98 89.50 81.00 17.47
#2 传统GGVF Snake优化文献[4]的结果 85.53 92.20 88.74 79.76 18.11
#2 文献[5]优化文献[4]的结果 88.60 88.06 88.33 79.10 17.52
#2 本文方法优化文献[4]的结果 87.02 96.41 91.47 84.29 19.65
#2 U-Net神经网络提取结果 97.98 92.29 95.05 90.57 17.33
#2 本文方法优化U-Net结果 97.82 93.00 95.35 91.12 38.61
#3 文献[8]的建筑物提取结果 81.89 97.34 88.95 80.10 9.14
#3 文献[4]的建筑物提取结果 79.77 99.31 88.48 79.34 8.90
#3 传统GGVF Snake优化文献[4]的结果 86.41 98.64 92.12 85.40 19.90
#3 文献[5]优化文献[4]的结果 91.36 92.83 92.09 85.34 20.76
#3 本文方法优化文献[4]的结果 98.46 94.72 96.55 93.34 38.06
#3 U-Net神经网络提取结果 82.56 96.78 89.10 80.35 12.05
#3 本文方法优化U-Net结果 86.78 96.96 91.58 84.48 20.01
表 1  对图3~5中不同方法结果的客观评价
图 6  利用不同方法对单个建筑物的优化结果对比
图 7  数据集的各项精度对比
方法 t /s
影像#1 影像#2 影像#3
传统GGVF Snake 35.21 548.32 465.34
文献[5]方法 3.92 216.45 145.08
本文方法 3.35 75.56 68.45
表 2  3种优化方法的运行效率对比
1 游永发, 王思远, 王斌, 等 高分辨率遥感影像建筑物分级提取[J]. 遥感学报, 2019, 23 (1): 125- 136
YOU Yong-fa, WANG Si-yuan, WANG Bin, et al Study on hierarchical building extraction from high resolution remote sensing imagery[J]. Journal of Remote Sensing, 2019, 23 (1): 125- 136
2 胡荣明, 黄小兵, 黄远程 增强形态学建筑物指数应用于高分辨率遥感影像中建筑物提取[J]. 测绘学报, 2014, 43 (5): 514- 520
HU Rong-ming, HUANG Xiao-bing, HUANG Yuan-cheng An enhanced morphological building index for building extraction from high-resolution images[J]. Acta Geodaetica et Cartographica Sinica, 2014, 43 (5): 514- 520
3 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
LIN Xiang-guo, ZHANG Ji-xian 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
doi: 10.11947/j.AGCS.2017.20170068
4 高贤君, 郑学东, 沈大江, 等 城郊高分影像中利用阴影的建筑物自动提取[J]. 武汉大学学报: 信息科学版, 2017, 42 (10): 1350- 1357
GAO Xian-jun, ZHENG Xue-dong, SHEN Da-jiang, et al Automatic building extraction based on shadow analysis from high resolution images in suburb areas[J]. Geomatics and Information Science of Wuhan University, 2017, 42 (10): 1350- 1357
5 常京新, 王双喜, 杨元维, 等 高分遥感影像建筑物轮廓的逐级优化方法[J]. 中国激光, 2020, 47 (10): 1010002
CHANG Jing-xin, WANG Shuang-xi, YANG Yuan-wei, et al Hierarchical optimization method of building contour in high-resolution remote sensing images[J]. Chinese Journal of Lasers, 2020, 47 (10): 1010002
doi: 10.3788/CJL202047.1010002
6 丁亚洲, 冯发杰, 吏军平, 等 多星形约束图割与轮廓规则化的高分遥感影像直角建筑物提取[J]. 测绘学报, 2018, 47 (12): 1630- 1639
DING Ya-zhou, FENG Fa-jie, LI Jun-ping, et al Right-angle buildings extraction from high-resolution aerial image based on multi-stars constraint segmentation and regularization[J]. Acta Geodaetica et Cartographica Sinica, 2018, 47 (12): 1630- 1639
doi: 10.11947/j.AGCS.2018.20170486
7 周绍光, 孙金彦, 凡莉, 等 高分辨率遥感影像的建筑物轮廓信息提取方法[J]. 国土资源遥感, 2015, 27 (3): 52- 58
ZHOU Shao-guang, SUN Jin-yan, FAN Li, et al Extraction of building contour from high resolution images[J]. Remote Sensing for Land and Resources, 2015, 27 (3): 52- 58
8 尹峰, 祁琼, 许博文 基于角点的高分辨率遥感影像建筑物提取[J]. 地理空间信息, 2018, 16 (10): 58- 61
YIN Feng, QI Qiong, XU Bo-wen Building extraction from high-resolution remote sensing images based on junction detection[J]. Geospatial Information, 2018, 16 (10): 58- 61
doi: 10.3969/j.issn.1672-4623.2018.10.017
9 XU C Y, PRINCE J L Generalized gradient vector flow external forces for active contours[J]. Signal Processing, 1998, 71 (2): 131- 139
doi: 10.1016/S0165-1684(98)00140-6
10 ZHANG G N, WANG W X, LANG F N, et al White blood cell extraction on fractional calculus and gradient vector flow snake for Leukocyte classification on support vector machines[J]. Journal of Medical Imaging and Health Informatics, 2018, 8 (6): 1249- 1257
doi: 10.1166/jmihi.2018.2425
11 袁胜古, 潘俊, 应荷香, 等 一种改进的GVF蛇模型水域矢量边界更新方法[J]. 武汉大学学报: 信息科学版, 2015, 40 (4): 503- 509
YUAN Sheng-gu, PAN Jun, YING He-xiang, et al A method of water vector edge updating based on improved GVF snake model[J]. Geomatics and Information Science of Wuhan University, 2015, 40 (4): 503- 509
12 王峰萍, 王卫星, 薛柏玉, 等 GVF Snake与显著特征相结合的高分辨率遥感图像道路提取[J]. 测绘学报, 2017, 46 (12): 1978- 1985
WANG Feng-ping, WANG Wei-xing, XUE Bai-yu, et al Road extraction from high-spatial-resolution remote sensing image by combining GVF snake with salient features[J]. Acta Geodaetica et Cartographica Sinica, 2017, 46 (12): 1978- 1985
doi: 10.11947/j.AGCS.2017.20170393
13 马铃, 王翠花, 魏玉, 等 Snake模型在图像分割中的应用研究[J]. 科技信息, 2013, 432 (4): 47- 48
MA Ling, WANG Cui-hua, WEI Yu, et al Application research of snake model in image segmentation[J]. Science and Technology Information, 2013, 432 (4): 47- 48
doi: 10.3969/j.issn.1001-9960.2013.04.039
14 郑健, 王继, 宋世铭 Canny双阈值算子在边缘提取中的优势[J]. 地理空间信息, 2019, 17 (11): 128- 130
ZHENG Jian, WANG Ji, SONG Shi-ming Advantages of Canny double threshold operator in edge extraction[J]. Geospatial Information, 2019, 17 (11): 128- 130
doi: 10.3969/j.issn.1672-4623.2019.11.036
15 辛超, 刘扬 基于概率霍夫变换的车道线识别算法[J]. 测绘通报, 2019, (Supple.2): 52- 55
XIN Chao, LIU Yang Research on lane recognition algorithm based on probability Hough transform[J]. Bulletin of Surveying and Mapping, 2019, (Supple.2): 52- 55
16 王征, 章品正, 梁晓云, 等 基于HSV颜色模型的广义梯度矢量流图像分割方法[J]. 数据采集与处理, 2005, 20 (4): 394- 397
WANG Zheng, ZHANG Pin-zheng, LIANG Xiao-yun, et al GGVF method for image segmentation based on HSV[J]. Journal of Data Acquisition and Processing, 2005, 20 (4): 394- 397
doi: 10.3969/j.issn.1004-9037.2005.04.006
17 SUZUKI S, BE K Topological structural analysis of digitized binary images by border following[J]. Computer Vision, Graphics, and Image Processing, 1985, 30 (1): 32- 46
doi: 10.1016/0734-189X(85)90016-7
18 ANAND A, TRIPATHY S S, KUMAR R S. An improved edge detection using morphological Laplacian of Gaussian operator [C]// Proceedings of the 2nd International Conference on signal Processing and Integrated Networks. Noida, India: IEEE, 2015: 532-536.
19 XU C Y, PRINCE J L Snakes, shapes, and gradient vector flow[J]. IEEE Transactions on Image Processing, 1998, 7 (3): 359- 369
doi: 10.1109/83.661186
20 OLAF R, PHILIPP F, THOMAS B U-Net: convolutional networks for biomedical image segmentation[J]. Medical Image Computing and Computer-Assisted Intervention, 2015, 9351: 234- 241
21 王双喜, 杨元维, 常京新, 等 高分辨率影像分类提取建筑物轮廓的优化方法[J]. 激光与光电子学进展, 2020, 57 (2): 022801
WANG Shuang-xi, YANG Yuan-wei, CHANG Jing-xin, et al Optimization of building contours by classifying high-resolution images[J]. Laser and Optoelectronics Progress, 2020, 57 (2): 022801
doi: 10.3788/LOP57.022801
[1] 冉树浩,胡玉龙,杨元维,高贤君,李熙,陈明珠. 基于样本形态变换的高分遥感影像建筑物提取[J]. 浙江大学学报(工学版), 2020, 54(5): 996-1006.
[2] 弓晓虹, 郑音飞, 秦佳乐, 周浩. 基于乘性梯度的医学超声图像边缘检测算法[J]. 浙江大学学报(工学版), 2014, 48(10): 1871-1878.
[3] 刘士荣, 王凯, 邱雪娜. 基于自适应混合高斯模型全方位视觉目标检测[J]. J4, 2010, 44(7): 1387-1393.
[4] 周泓, 徐海儿, 耿晨歌. 基于HSI模型和Hough变换的指针式汽车仪表自动校验[J]. J4, 2010, 44(6): 1108-1112.
[5] 许晓燕, 夏顺仁, LIU Tian-ming, 等. 基于Hough变换的斑马鱼胚胎图像分析技术[J]. J4, 2009, 43(11): 2034-2037.
[6] 屈稳太. 基于弦中点Hough变换的椭圆检测方法[J]. J4, 2005, 39(8): 1132-1135.
[7] 陈慧芳 严惠民 姚晓强. 被动测距系统汽车目标提取算法研究[J]. J4, 2005, 39(4): 526-529.