多特征混合核SVM模型的遥感影像变化检测
|
摘要
针对传统变化检测方法会存在明显的"椒盐现象"以及不同核函数对同一特征性能表现差别比较大的问题,借鉴面向对象思想,提出多特征混合核支持向量机(support vector machine,SVM)模型的变化检测方法。首先,依据高空间分辨率遥感影像对象不同特征的变化检测优势,提取影像多种特征;然后,利用多种特征的多核函数组合,给出多特征混合核函数的构造方法;最后,构建基于多特征混合核SVM的变化检测模型,充分挖掘变化目标的完整性与准确性。实验结果表明,该方法能综合利用多种特征信息,检测精度明显高于单一特征,有利于提取小样本的变化信息,避免了以往检测方法需要确定变化阈值的复杂性和不确定性。
In view of the fact that different kernel functions have greatly different performance on the same feature,the authors propose a new method of change detection of multi-feature hybrid kernel support vector machine(SVM) model.According to the different characteristics of the change detection,the authors extract image features,make use of the multi-kernel function of several features,give the methods of constructing multi-feature and mixed-kernel function,construct change detection model of multi-feature mixed-nuclear support vector machine,and fully tap the integrity and accuracy of the varying target.The experimental results show that this method makes use of the information of various features.The detection precision is obviously higher than that of the single feature.The method not only takes advantage of extracting change information of small samples,but also avoids the complexity and uncertainty of the old detection method for determining the change threshold.
In view of the fact that different kernel functions have greatly different performance on the same feature,the authors propose a new method of change detection of multi-feature hybrid kernel support vector machine(SVM) model.According to the different characteristics of the change detection,the authors extract image features,make use of the multi-kernel function of several features,give the methods of constructing multi-feature and mixed-kernel function,construct change detection model of multi-feature mixed-nuclear support vector machine,and fully tap the integrity and accuracy of the varying target.The experimental results show that this method makes use of the information of various features.The detection precision is obviously higher than that of the single feature.The method not only takes advantage of extracting change information of small samples,but also avoids the complexity and uncertainty of the old detection method for determining the change threshold.
引文
[1]夏朝旭,何政伟,于欢,等.面向对象的土地覆被变化检测研究[J].遥感技术与应用,2014,29(1):106-113.Xia C X,He Z W,Yu H,et al.Study on land cover change detection based on object-oriented[J].Remote Sensing Technology and Application,2014,29(1):106-113.
[2]Zhang G,Li Y,Li Z J.A new approach toward object-based change detection[J].Science China Technological Sciences,2010,53(s1):105-110.
[3]Li F,Fang F M,Zhang G X.Unsupervised change detection in SARimages using curvelet and L1-norm based soft segmentation[J].International Journal of Remote Sensing,2016,37(14):3232-3254.
[4]Bovolo F,Bruzzone L,Marconcini M.A novel approach to unsupervised change detection based on a semisupervised SVM and a similarity measure[J].IEEE Transactions on Geoscience and Remote Sensing,2008,46(7):2070-2082.
[5]庄会富,邓喀中,范洪冬.纹理特征向量与最大化熵法相结合的SAR影像非监督变化检测[J].测绘学报,2016,45(3):339-346.Zhuang H F,Deng K Z,Fan H D.SAR images unsupervised change detection based on combination of texture feature vector with maximum entropy principle[J].Acta Geodaetica et Cartographica Sinica,2016,45(3):339-346.
[6]钟家强,王润生.基于自适应参数估计的多时相遥感图像变化检测[J].测绘学报,2005,34(4):331-336.Zhong J Q,Wang R S.Multitemporal remote sensing image change detection based on adaptive parameter estimation[J].Acta Geodaetica et Cartographica Sinica,2005,34(4):331-336.
[7]魏立飞,钟燕飞,张良培,等,遥感影像融合的自适应变化检测[J].遥感学报,2010,14(6):1196-1211.Wei L F,Zhong Y F,Zhang L P,et al.Adaptive change method of remote sensing image fusion[J].Journal of Remote Sensing,2010,14(6):1196-1211.
[8]林怡,刘冰,陈映鹰,等,多特征差分核支持向量机遥感影像变化检测方法[J].武汉大学学报(信息科学版),2013,38(8):978-982.Lin Y,Liu B,Chen Y Y,et al.Change detection method based on multi-feature differencing kernel SVM for remote sensing imagery[J].Geomatics and Information Science of Wuhan University,2013,38(8):978-982.
[9]Zien A,Ong C S.Multiclass multiple kernel learning[C]//Proceedings of the Twenty-Fourth International Conference Machine Learning(ICML).Corvallis:ACM,2007.
[10]谭琨,杜培军.基于支持向量机的高光谱遥感图像分类[J].红外与毫米波学报,2008,27(2):123-128.Tang K,Du P J.Hyperspectral remote sensing image classification based on support vector machine[J].Journal of Information and Millimeter Waves,2008,27(2):123-128.
[11]田淞,宋建社,张雄美,等.KM-SVM法的SAR图像无监督变化检测[J].系统工程与电子技术,2015,37(5):1042-1046.Tian S,Song J S,Zhang X M,et al.KM-SVM approach to unsupervised change detection in SAR images[J].Systems Engineering and Electronics,2015,37(5):1042-1046.
[12]夏晨阳,石爱业,吴国宝.基于SVM混合核的遥感图像变化检测[J].信息技术,2014(8):38-41.Xia C Y,Shi A Y,Wu G B.Remote sensing image change detection based on mixed kernel of SVM[J].Information Technology,2014(8):38-41.
[13]祁增营,王京,左正立.湿地变化监测研究现状与展望[J].遥感信息,2012,27(6):124-132.Qi Z Y,Wang J,Zuo Z L.Current status and prospect of researches on wetland change monitoring[J].Remote Sensing Information,2012,27(6):124-132.
[14]慕彩红,霍利利,刘逸,等,基于小波融合和PCA-核模糊聚类的遥感图像变化检测[J].电子学报,2015,43(7):1375-1381.Mu C H,Huo L L,Liu Y,et al.Change detection for remote sensing images based on wavelet fusion and PCA-kernel fuzzy clustering[J].Acta Electronica Sinica,2015,43(7):1375-1381.
[15]Wang X G,Jia Z H,Yang J,et al.Change detection in SAR images based on the logarithmic transformation and total variation denoising method[J].Remote Sensing Letters,2017,8(3):214-223.
[16]王光辉,李建磊,王华斌,等.基于多特征融合的遥感影像变化检测算法[J].国土资源遥感,2018,30(2):93-99.doi:10.6046/gtzyyg.2018.02.13.Wang G H,Li J L,Wang H B,et al.Change detection based on adaptive fusion of multiple features[J].Remote Sensing for Land and Resources,2018,30(2):93-99.doi:10.6046/gtzyyg.2018.02.13.
[17]尹凌宇,覃先林,孙桂芬,等.利用KPCA法检测高分一号影像中的森林覆盖变化[J].国土资源遥感,2018,30(1):95-101.doi:10.6046/gtzyyg.2018.01.13.Yin L Y,Qin X L,Sun G F,et al.The method for detecting forest cover change in GF-1images by using KPCA[J].Remote Sensing for Land and Resources,2018,30(1):95-101.doi:10.6046/gtzyyg.2018.01.13.
[18]邓曾,李丹,柯樱海,等.基于改进SVM算法的高分辨率遥感影像分类[J].国土资源遥感,2016,28(3):12-18.doi:10.6046/gtzyyg.2016.03.03.Deng Z,Li D,Ke Y H,et al.An improved SVM algorithm for high spatial resolution remote sensing image classification[J].Remote Sensing for Land and Resources,2016,28(3):12-18.doi:10.6046/gtzyyg.2016.03.03.
[2]Zhang G,Li Y,Li Z J.A new approach toward object-based change detection[J].Science China Technological Sciences,2010,53(s1):105-110.
[3]Li F,Fang F M,Zhang G X.Unsupervised change detection in SARimages using curvelet and L1-norm based soft segmentation[J].International Journal of Remote Sensing,2016,37(14):3232-3254.
[4]Bovolo F,Bruzzone L,Marconcini M.A novel approach to unsupervised change detection based on a semisupervised SVM and a similarity measure[J].IEEE Transactions on Geoscience and Remote Sensing,2008,46(7):2070-2082.
[5]庄会富,邓喀中,范洪冬.纹理特征向量与最大化熵法相结合的SAR影像非监督变化检测[J].测绘学报,2016,45(3):339-346.Zhuang H F,Deng K Z,Fan H D.SAR images unsupervised change detection based on combination of texture feature vector with maximum entropy principle[J].Acta Geodaetica et Cartographica Sinica,2016,45(3):339-346.
[6]钟家强,王润生.基于自适应参数估计的多时相遥感图像变化检测[J].测绘学报,2005,34(4):331-336.Zhong J Q,Wang R S.Multitemporal remote sensing image change detection based on adaptive parameter estimation[J].Acta Geodaetica et Cartographica Sinica,2005,34(4):331-336.
[7]魏立飞,钟燕飞,张良培,等,遥感影像融合的自适应变化检测[J].遥感学报,2010,14(6):1196-1211.Wei L F,Zhong Y F,Zhang L P,et al.Adaptive change method of remote sensing image fusion[J].Journal of Remote Sensing,2010,14(6):1196-1211.
[8]林怡,刘冰,陈映鹰,等,多特征差分核支持向量机遥感影像变化检测方法[J].武汉大学学报(信息科学版),2013,38(8):978-982.Lin Y,Liu B,Chen Y Y,et al.Change detection method based on multi-feature differencing kernel SVM for remote sensing imagery[J].Geomatics and Information Science of Wuhan University,2013,38(8):978-982.
[9]Zien A,Ong C S.Multiclass multiple kernel learning[C]//Proceedings of the Twenty-Fourth International Conference Machine Learning(ICML).Corvallis:ACM,2007.
[10]谭琨,杜培军.基于支持向量机的高光谱遥感图像分类[J].红外与毫米波学报,2008,27(2):123-128.Tang K,Du P J.Hyperspectral remote sensing image classification based on support vector machine[J].Journal of Information and Millimeter Waves,2008,27(2):123-128.
[11]田淞,宋建社,张雄美,等.KM-SVM法的SAR图像无监督变化检测[J].系统工程与电子技术,2015,37(5):1042-1046.Tian S,Song J S,Zhang X M,et al.KM-SVM approach to unsupervised change detection in SAR images[J].Systems Engineering and Electronics,2015,37(5):1042-1046.
[12]夏晨阳,石爱业,吴国宝.基于SVM混合核的遥感图像变化检测[J].信息技术,2014(8):38-41.Xia C Y,Shi A Y,Wu G B.Remote sensing image change detection based on mixed kernel of SVM[J].Information Technology,2014(8):38-41.
[13]祁增营,王京,左正立.湿地变化监测研究现状与展望[J].遥感信息,2012,27(6):124-132.Qi Z Y,Wang J,Zuo Z L.Current status and prospect of researches on wetland change monitoring[J].Remote Sensing Information,2012,27(6):124-132.
[14]慕彩红,霍利利,刘逸,等,基于小波融合和PCA-核模糊聚类的遥感图像变化检测[J].电子学报,2015,43(7):1375-1381.Mu C H,Huo L L,Liu Y,et al.Change detection for remote sensing images based on wavelet fusion and PCA-kernel fuzzy clustering[J].Acta Electronica Sinica,2015,43(7):1375-1381.
[15]Wang X G,Jia Z H,Yang J,et al.Change detection in SAR images based on the logarithmic transformation and total variation denoising method[J].Remote Sensing Letters,2017,8(3):214-223.
[16]王光辉,李建磊,王华斌,等.基于多特征融合的遥感影像变化检测算法[J].国土资源遥感,2018,30(2):93-99.doi:10.6046/gtzyyg.2018.02.13.Wang G H,Li J L,Wang H B,et al.Change detection based on adaptive fusion of multiple features[J].Remote Sensing for Land and Resources,2018,30(2):93-99.doi:10.6046/gtzyyg.2018.02.13.
[17]尹凌宇,覃先林,孙桂芬,等.利用KPCA法检测高分一号影像中的森林覆盖变化[J].国土资源遥感,2018,30(1):95-101.doi:10.6046/gtzyyg.2018.01.13.Yin L Y,Qin X L,Sun G F,et al.The method for detecting forest cover change in GF-1images by using KPCA[J].Remote Sensing for Land and Resources,2018,30(1):95-101.doi:10.6046/gtzyyg.2018.01.13.
[18]邓曾,李丹,柯樱海,等.基于改进SVM算法的高分辨率遥感影像分类[J].国土资源遥感,2016,28(3):12-18.doi:10.6046/gtzyyg.2016.03.03.Deng Z,Li D,Ke Y H,et al.An improved SVM algorithm for high spatial resolution remote sensing image classification[J].Remote Sensing for Land and Resources,2016,28(3):12-18.doi:10.6046/gtzyyg.2016.03.03.