联合像素与多尺度对象的高分辨率遥感影像谱聚类分割

doi:10.13229/j.cnki.jdxbgxb20180537

摘要/Abstract

摘要：

建立了融合多尺度信息的图模型，同时，为了顾及对象的局部统计特性，改进了基于对象的顶点间相似度计算方法。在图模型的基础上完成相似矩阵计算，并使用归一化分割准则对相似矩阵特征进行分解，将原始数据映射到低维子空间。最后，对特征筛选后的子集使用聚类算法完成影像分割。为了验证本文方法的有效性，选取高空间分辨率遥感影像进行实验并与目前分割精度较高的算法做定量化对比。实验结果表明：在4项实验指标中，除一项基本持平外，其他3项指标优于其他算法，证明了本文方法在高分辨率遥感影像分割中的有效性。

关键词: 摄影测量与遥感, 遥感影像分割, 谱聚类, 归一化分割, 多尺度对象

Abstract:

An spectral clustering segmentation method for high spatial resolution (HSR) remote sensing image based on combination of pixels and multi-scale objects is proposed. The algorithm first focus on building a graph that integrates multi-scale information as well as improve similarity computation method between objects. Then, the similarity matrix is computed on the graph, and normalized cut criterion is used to similarity matrix eigen-decomposition so that the original data is mapped into the low-dimensional subspace. Finally, the clustering algorithm is used to complete the image segmentation after the selected subset of the eigenvectors. In order to prove the effectiveness of the algorithm, we choose high spatial resolution remote sensing images to conduct experiment and compare to the state-of-the-art techniques. The experimental result show that three of four experimental indexes are superior to other algorithms, which proves the effectiveness of this proposed method segmentation method.

Key words: photogrammetry and remote sensing, remote sensing image segmentation, spectral clustering, normalized cut, multi?scale objects

中图分类号:

李军军,曹建农,程贝贝,廖娟,朱莹莹. 联合像素与多尺度对象的高分辨率遥感影像谱聚类分割[J]. 吉林大学学报(工学版), 2019, 49(6): 2098-2108.

Jun-jun LI,Jian-nong CAO,Bei-bei CHENG,Juan LIAO,Ying-ying ZHU. High spatial resolution remote sensing imagery segmentation based on combination of pixels and multi⁃scaleobjects using spectral clustering[J]. Journal of Jilin University(Engineering and Technology Edition), 2019, 49(6): 2098-2108.

图/表 7

图1

图2

图3

表1

表2

图4

图5

参考文献

1	Moser G , Serpico S B , Benediktsson J A . Land-Cover mapping by markov modeling of spatial–contextual information in very-high-resolution remote sensing images[J]. Proceedings of the IEEE,2013,101(3):631-651.
2	Liu H , Zhao F , Jiao L . Fuzzy spectral clustering with robust spatial information for image segmentation[J]. Applied Soft Computing,2012,12(11):3636-3647.
3	Ng A Y , Jordan M I , Weiss Y . On spectral clustering: analysis and an algorithm[C]∥Proceedings of the 14th International Conference on Neural Information Processing Systems: Natural and Synthetic,Vancouver, Canada,2001:849-856.
4	Luxburg U V . A tutorial on spectral clustering[J]. Statistics and Computing,2007,17(4):395-416.
5	Shi J B , Malik J . Normalized cuts and image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(8):888-905.
6	袁永华，李玉，赵雪梅 . 基于谱聚类的高分辨率全色遥感影像分割[J]. 仪器仪表学报,2016,37(7): 1656-1664.
6	Yuan Yong-hua , Li Yu , Zhao Xue-mei . High-resolution panchromatic remote sensing image segmentation based on spectral clustering[J]. Chinese Journal of Scientific Instrument,2016,37(7):1656-1664.
7	张向荣，骞晓雪，焦李成 . 基于免疫谱聚类的图像分割[J]. 软件学报,2010,21(9):2196-2205.
7	Zhang Xiang-rong , Qian Xiao-xue , Jiao Li-cheng . Immune spectral clustering algorithm for image segmentation[J]. Journal of Software,2010,21(9): 2196-2205.
8	董安国，李佳逊，张蓓，等 . 基于谱聚类和稀疏表示的高光谱图像分类算法[J]. 光学学报,2017,37(8): 363-370.
8	Dong An-guo , Li Jia-xun , Zhang Bei ,et al . Hyperspectral image classification algorithm based on spectral clustering and sparse representation[J]. Acta Optica Sinica,2017,37(8):363-370.
9	Cour T , Benezit F , Shi J . Spectral segmentation with multiscale graph decomposition[C]∥2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Diego, USA, 2005:1124-1131.
10	刘仲民,李战明,李博皓,等 . 基于稀疏矩阵的谱聚类图像分割算法[J].吉林大学学报:工学版,2017,47(4):1308-1313.
10	Liu Zhong-min , Li Zhan-ming , Li Bo-hao , et al . Spectral clustering image segmentation based on sparse matrix[J]. Journal of Jilin University(Engineering and Technology Edition),2017,47(4):1308-1313.
11	Gui Y , Bai X , Li Z , et al . Color image segmentation using mean shift and improved spectral clustering[C]∥2012 12th International Conference on Control Automation Robotics & Vision, Guangzhou, China,2012:1386-1391.
12	Cheng H D , Jiang X H , Sun Y , et al . Color image segmentation: advances and prospects[J]. Pattern Recognition,2001,34(12):2259-2281.
13	Comaniciu D , Meer P . Mean shift: a robust approach toward feature space analysis[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(5):603-619.
14	宋熙煜,周利莉,李中国,等 . 图像分割中的超像素方法研究综述[J]. 中国图象图形学报,2015,20(5): 599-608.
14	Song Xi-yu , Zhou Li-li , Li Zhong-guo , et al . Review on superpixel methods in image segmentation[J]. Journal of Image and Graphics,2015,20(5):599-608.
15	Zha H Y , He X F , Ding C , et al . Bipartite graph partitioning and data clustering[DB/OL].[2018-05-12].https:∥arxiv.org/pdf/cs/0108018.pdf.
16	Dhillon I S . Co-clustering documents and words using bipartite spectral graph partitioning[C]∥Proceedings of the Seventh ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, California, 2001: 269-274.
17	Li Z , Wu X M , Chang S F . Segmentation using superpixels: a bipartite graph partitioning approach[C]∥2012 IEEE Conference on Computer Vision and Pattern Recognition, Providence, USA, 2012:789-796.
18	Kim T H , Lee K M , Lee S U . Learning full pairwise affinities for spectral segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2013,35(7):1690-1703.
19	Kohli P , Ladicky L , Torr P H S . Robust higher order potentials for enforcing label consistency[J/OL]. [2018-05-16].http:∥.
20	Takamura S , Kobayashi N . Practical extension to CIELUV color space to improve uniformity[J]. Proceedings International Conference on Image Processing,2002,2(2):393-396.
21	Zhong Yan-fei , Gao Rong-rong , Zhang Liang-pei . Multiscale and multifeature normalized cut segmentation for high spatial resolution remote sensing imagery[J]. IEEE Transactions on Geoscience and Remote Sensing,2016,54(10):6061-6075.
22	An V , Oraintara S . A study of relative phase in complex wavelet domain: property, statistics and applications in texture image retrieval and segmentation[J]. Signal Processing Image Communication,2010,25(1):28-46.
23	Wu Z , Leahy R . An optimal graph theoretic approach to data clustering: theory and its application to image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1993,15(11):1101-1113.
24	Hagen L , Kahng A B . New spectral methods for ratio cut partitioning and clustering[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems,2002,11(9):1074-1085.
25	Pothen A , Simon H D , Liou K P . Partitioning sparse matrices with eigenvectors of graphs[J]. SIAM Journal on Matrix Analysis and Applications,1990,11(3):430-452.
26	Yu S X , Shi J . Multiclass spectral clustering[C]∥Proceedings Ninth IEEE International Conference on Computer Vision, Nice, France, 2003: 313-319.
27	Unnikrishnan R , Pantofaru C , Hebert M . Toward objective evaluation of image segmentation algorithms[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence,2007,29(6):929-944.
28	Martin D , Fowlkes C , Tal D , et al . A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics[C]∥Eighth IEEE International Conference on Computer Vision, Canada,2001:37-42.
29	Marina M . Comparing clusterings: an axiomatic view[C]∥International Conference, Germany,2005:577-584.
30	Freixenet J , Mu?oz X , Raba D , et al . Yet another survey on image segmentation: region and boundary information integration[C]∥7th European Conference on Computer Vision, Copenhagen, Denmark, 2002: 408-422.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

指标	影像1	影像2	影像3	影像4	影像5	影像6	影像7	影像8	影像9	影像10
PRI	0.905 7	0.967 00	0.975 1	0.972 4	0.975 6	0.943 0	0.833 0	0.952 5	0.967 1	0.958 0
GCE	2.340 9	1.564 10	2.153 5	2.238 8	1.976 0	1.926 0	2.269 2	2.267 2	1.473 0	1.825 7
VoI	0.272 3	0.234 90	0.333 0	0.351 1	0.302 4	0.273 2	0.230 3	0.312 1	0.207 7	0.250 4
BDE	6.371 3	2.431 90	5.868 7	2.629 0	3.284 1	4.342 8	4.211 1	4.955 6	2.370 0	3.263 2

指标	影像1	影像2	影像3	影像4	影像5	影像6	影像7	影像8	影像9	影像10
PRI	0.903 4	0.964 4	0.979 5	0.977 7	0.966 2	0.943 6	0.837 7	0.961 5	0.969 5	0.936 8
GCE	2.527 5	1.688 1	2.203 2	2.284 3	1.900 7	1.968 5	2.395 8	2.330 2	1.500 1	2.286 0
VoI	0.325 4	0.239 0	0.357 1	0.380 8	0.305 5	0.312 0	0.296 0	0.352 2	0.214 1	0.354 6
BDE	7.581 51	2.701 3	6.107 2	2.508 1	3.271 2	4.786 3	4.661 5	4.874 3	2.367 3	3.793 9

[1]	邓剑勋, 熊忠阳, 邓欣. 基于谱聚类矩阵的改进DNALA算法[J]. 吉林大学学报(工学版), 2018, 48(3): 903-908.
[2]	刘仲民, 李战明, 李博皓, 胡文瑾. 基于稀疏矩阵的谱聚类图像分割算法[J]. 吉林大学学报(工学版), 2017, 47(4): 1308-1313.
[3]	李晖晖, 滑立, 杨宁, 刘坤. 基于MSA特征和模拟退火优化的遥感图像多目标关联算法[J]. 吉林大学学报(工学版), 2015, 45(4): 1353-1359.
[4]	姜超, 耿则勋, 刘立勇, 潘映峰. 结合噪声去除的极大似然图像复原[J]. 吉林大学学报(工学版), 2015, 45(4): 1360-1366.
[5]	王京萌, 张爱武, 赵宁宁, 孟宪刚. 斜采样的倾斜角度对采样产生混叠的影响及其与分辨率的关系[J]. 吉林大学学报(工学版), 2015, 45(3): 953-960.
[6]	曹建农, 郭佳, 王蓓, 董昱威, 王平禄. 高分辨率影像中城区树冠多尺度聚类识别方法[J]. 吉林大学学报(工学版), 2014, 44(4): 1215-1224.
[7]	曲琳,周凡,陈耀武. 基于Hausdorff距离的视觉监控轨迹分类算法[J]. 吉林大学学报(工学版), 2009, 39(06): 1618-1624.