An intelligent classification method for building shape based on fusion of global and local features

doi:10.11947/j.AGCS.2024.20240040

Abstract

Abstract:

Supported by deep learning methods for building shape cognition, it has become a hot research topic in fields such as cartography. The feature mining ability of deep learning can help extract embedded representations of shapes, supporting application scenarios such as cartographic generalization and spatial retrieval. A graph convolutional neural network model for building shape classification that integrates global features and graph node features is constructed, and validated using building data as an example. Firstly, a weighted building graph is constructed, and then a fusion description of the shape is generated based on the 4 macroscopic shape features of building and the multi-level local and regional structural features of boundary vertice. Graph convolutional neural networks are used to extract multi-level shape information, and the feature coding generated by fusing graph representations from different layers is used for shape classification.The experimental results show that compared to the comparative method, the proposed method is more effective in distinguishing the shape categories of different buildings, and the generated feature coding have positive shape discrimination.

Key words: shape cognition, graph convolutional neural network, building shape classification, feature fusion, graph classification

CLC Number:

P208

Fubing ZHANG, Qun SUN, Jingzhen MA, Shijie SUN, Bowei WEN. An intelligent classification method for building shape based on fusion of global and local features[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(9): 1842-1852.

Figures/Tables 15

Fig.1

Fig.2

Fig.3

Fig.4

Tab.1

Fig.5

Fig.6

Fig.7

Fig.8

Tab.2

Tab.3

Tab.4

Tab.5

Fig.9

Tab.6

References 31

[1]	晏雄锋, 艾廷华, 杨敏, 等. 地图空间形状认知的自编码器深度学习方法[J]. 测绘学报, 2021, 50(6):757-765. DOI: 10.11947/j.AGCS.2021.20210046.
	YAN Xiongfeng, AI Tinghua, YANG Min, et al. Shape cognition in map space using deep auto-encoder learning[J]. Acta Geodaetica et Cartographica Sinica, 2021, 50(6):757-765. DOI: 10.11947/j.AGCS.2021.20210046.
[2]	艾廷华, 帅赟, 李精忠. 基于形状相似性识别的空间查询[J]. 测绘学报, 2009, 38(4):356-362.
	AI Tinghua, SHUAI Yun, LI Jingzhong. A spatial query based on shape similarity cognition[J]. Acta Geodaetica et Cartographica Sinica, 2009, 38(4):356-362.
[3]	YAN Xiongfeng, AI Tinghua, YANG Min, et al. Graph convolutional autoencoder model for the shape coding and cognition of buildings in maps[J]. International Journal of Geographical Information Science, 2021, 35(3):490-512.
[4]	王斌. 一种基于多级弦长函数的傅立叶形状描述子[J]. 计算机学报, 2010, 33(12):2387-2396.
	WANG Bin. A Fourier shape descriptor based on multi-level chord length function[J]. Chinese Journal of Computers, 2010, 33(12):2387-2396.
[5]	BRODERSEN L, ANDERSEN H H K, WEBER S. Applying eye-movement tracking for the study of map perception and map design[R]. Copenhagen: National Survey and Cadastre Publications, 2002.
[6]	晏雄锋, 艾廷华, 杨敏. 居民地要素化简的形状识别与模板匹配方法[J]. 测绘学报, 2016, 45(7):874-882. DOI: 10.11947/j.AGCS.2016.20150162.
	YAN Xiongfeng, AI Tinghua, YANG Min. A simplification of residential feature by the shape cognition and template matching method[J]. Acta Geodaetica et Cartographica Sinica, 2016, 45(7):874-882. DOI: 10.11947/j.AGCS.2016.20150162.
[7]	晏雄锋, 袁拓, 杨敏, 等. 建筑物形状特征分析表达与自适应化简方法[J]. 测绘学报, 2022, 51(2):269-278. DOI: 10.11947/j.AGCS.2022.20210302.
	YAN Xiongfeng, YUAN Tuo, YANG Min, et al. An adaptive building simplification approach based on shape analysis and representation[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(2):269-278. DOI: 10.11947/j.AGCS.2022.20210302.
[8]	YAN Xiongfeng, AI Tinghua, YANG Min, et al. A graph convolutional neural network for classification of building patterns using spatial vector data[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2019, 150:259-273.
[9]	LI Wenwen, GOODCHILD M F, CHURCH R. An efficient measure of compactness for two-dimensional shapes and its application in regionalization problems[J]. International Journal of Geographical Information Science, 2013, 27(6):1227-1250.
[10]	BASARANER M, CETINKAYA S. Performance of shape indices and classification schemes for characterising perceptual shape complexity of building footprints in GIS[J]. International Journal of Geographical Information Science, 2017, 31(10):1952-1977.
[11]	FREEMAN H. On the encoding of arbitrary geometric configurations[J]. IEEE Transactions on Electronic Computers, 1961, EC-10(2):260-268.
[12]	刘鹏程, 艾廷华, 胡晋山, 等. 基于原型模板形状匹配的建筑多边形化简[J]. 武汉大学学报(信息科学版), 2010, 35(11):1369-1372.
	LIU Pengcheng, AI Tinghua, HU Jinshan, et al. Building-polygon simplification based on shape matching of prototype template[J]. Geomatics and Information Science of Wuhan University, 2010, 35(11):1369-1372.
[13]	魏智威, 郭庆胜, 程璐, 等. 建筑物图形形状相似性计算的序列分析法[J]. 测绘学报, 2021, 50(12):1683-1693. DOI: 10.11947/j.AGCS.2021.20200227.
	WEI Zhiwei, GUO Qingsheng, CHENG Lu, et al. Shape similarity measurement based on DNA alignment for buildings with multiple orthogonal features[J]. Acta Geodaetica et Cartographica Sinica, 2021, 50(12):1683-1693. DOI: 10.11947/j.AGCS.2021.20200227.
[14]	ARKIN E M, CHEW L P, HUTTENLOCHER D P, et al. An efficiently computable metric for comparing polygonal shapes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1991, 13(3):209-216.
[15]	AI Tinghua, CHENG Xiaoqiang, LIU Pengcheng, et al. A shape analysis and template matching of building features by the Fourier transform method[J]. Computers, Environment and Urban Systems, 2013, 41:219-233.
[16]	BELONGIE S, MALIK J, PUZICHA J. Shape matching and object recognition using shape contexts[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, 24(4):509-522.
[17]	艾廷华. 深度学习赋能地图制图的若干思考[J]. 测绘学报, 2021, 50(9):1170-1182. DOI: 10.11947/j.AGCS.2021.20210091.
	AI Tinghua. Some thoughts on deep learning enabling cartography[J]. Acta Geodaetica et Cartographica Sinica, 2021, 50(9):1170-1182. DOI: 10.11947/j.AGCS.2021.20210091.
[18]	马磊, 闫浩文, 王中辉, 等. 机器自监督学习的建筑物面要素几何形状度量[J]. 测绘科学, 2017, 42(12):171-177.
	MA Lei, YAN Haowen, WANG Zhonghui, et al. Geometry shape measurement of building surface elements based on self-supervised machine learning[J]. Science of Surveying and Mapping, 2017, 42(12):171-177.
[19]	焦洋洋, 刘平芝, 刘爱龙, 等. AlexNet支持下的地图建筑物形状分类方法[J]. 地球信息科学学报, 2022, 24(12):2333-2341.
	JIAO Yangyang, LIU Pingzhi, LIU Ailong, et al. Map building shape classification method based on Alexnet[J]. Journal of Geo-information Science, 2022, 24(12):2333-2341.
[20]	HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8):1735-1780.
[21]	刘鹏程, 黄欣, 马宏然, 等. 建筑物多边形高精度识别的傅里叶形状描述子神经网络方法[J]. 测绘学报, 2022, 51(9):1969-1976. DOI: 10.11947/j.AGCS.2022.20210730.
	LIU Pengcheng, HUANG Xin, MA Hongran, et al. Fourier descriptor-based neural network method for high-precision shape recognition of building polygon[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(9):1969-1976. DOI: 10.11947/j.AGCS.2022.20210730.
[22]	NIEPERT M, AHMED M, KUTZKOV K, Learning convolutional neural networks for graphs[C]//Proceedings of the 33rd International Conference on Machine Learning. New York: Curran Associates, 2016: 2014-2023.
[23]	KIPF T N, WELLING M. Semi-supervised classification with graph convolutional networks[EB/OL]. [2024-05-30]. https://arxiv.org/abs/1609.02907v4.
[24]	于洋洋, 贺康杰, 武芳, 等. 面状居民地形状分类的图卷积神经网络方法[J]. 测绘学报, 2022, 51(11):2390-2402. DOI: 10.11947/j.AGCS.2022.20210134.
	YU Yangyang, HE Kangjie, WU Fang, et al. Graph convolution neural network method for shape classification of areal settlements[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(11):2390-2402. DOI: 10.11947/j.AGCS.2022.20210134.
[25]	YANG Chengzhuan, WEI Hui, YU Qian. A novel method for 2D nonrigid partial shape matching[J]. Neurocomputing, 2018, 275:1160-1176.
[26]	ZHANG Dengsheng, LU Guojun. Study and evaluation of different Fourier methods for image retrieval[J]. Image and Vision Computing, 2005, 23(1):33-49.
[27]	安晓亚, 孙群, 肖强, 等. 一种形状多级描述方法及在多尺度空间数据几何相似性度量中的应用[J]. 测绘学报, 2011, 40(4):495-501, 508.
	AN Xiaoya, SUN Qun, XIAO Qiang, et al. A shape multilevel description method and application in measuring geometry similarity of multi-scale spatial data[J]. Acta Geodaetica et Cartographica Sinica, 2011, 40(4):495-501, 508.
[28]	程绵绵, 孙群, 徐立, 等. 面轮廓线相似性和复杂性度量及在化简中的应用[J]. 测绘学报, 2019, 48(4):489-501. DOI: 10.11947/j.AGCS.2019.20180124.
	CHENG Mianmian, SUN Qun, XU Li, et al. Polygon contour similarity and complexity measurement and application in simplification[J]. Acta Geodaetica et Cartographica Sinica, 2019, 48(4):489-501. DOI: 10.11947/j.AGCS.2019.20180124.
[29]	HAMMOND D K, VANDERGHEYNST P, GRIBONVAL R. Wavelets on graphsvia spectral graph theory[J]. Applied and Computational Harmonic Analysis, 2011, 30(2):129-150.
[30]	LI Pengbo, YAN Haowen, LU Xiaomin. A siamese neural network for learning the similarity metrics of linear features[J]. International Journal of Geographical Information Science, 2023, 37(3):684-711.
[31]	VAN DER MAATEN L, HINTON G. Visualizing data using t-SNE[J]. Journal of Machine Learning Research, 2008, 9(1):2579-2605.

参考样本	K值	对比样本
参考样本	K值
	1	1.0	1.0	1.0	1.0	1.0
	2	1.0	1.0	1.0	1.0	1.0

参考样本	K值	对比样本
参考样本	K值
	1	0.978 7	0.920 6	0.983 4	0.943 3	0.932 5
	2	0.991 6	0.956 4	0.988 6	0.955 4	0.965 5

参考样本	K值	对比样本
参考样本	K值
	1	0.978 7	0.943 3	0.983 4	0.920 6	0.996 9
	2	0.991 6	0.955 4	0.988 6	0.956 4	0.998 3

方法	分类精度
ED	17.2
TA	79.5
GACE	94.1
NN	96.4
Spatial－GCN	92.3
Spectral－GCN(K=1)	97.6
Spectral－GCN(K=2)	98.1

检索结果	检索模板
检索结果
Top 1
Top 1	0.996 1	0.923 2	0.936 3	0.963 1	0.960 0	0.929 5
Top 2
Top 2	0.993 4	0.916 8	0.936 1	0.938 0	0.923 3	0.903 0
Bottom 1
Bottom 1	0.630 9	0.541 0	0.591 9	0.619 7	0.674 1	0.566 0
Bottom 2
Bottom 2	0.634 6	0.546 1	0.629 5	0.620 2	0.709 6	0.566 4