基于分层特征描述的舰船目标鉴别
|
摘要
针对当前一些目标鉴别方法无法兼顾目标的可分性和方法的有效性,同时又能减少计算的复杂度等要求,提出了一种基于分层特征描述的鉴别方法。首先,提取目标的简单形状或几何特征,利用加权投票法初步筛选并去除大量易识别的虚警;然后对筛选的候选目标提取更为复杂的鉴别特征,利用特征分离法选择最优特征组合,并采用支持向量机方法进行二次鉴别,进一步去除虚警,得到真实目标。实验结果表明,该方法对目标的整体检测效果较好,具有较高的可区分性和可鉴别性;能有效减少计算的复杂度,同时又能在一定程度上减少外界因素的影响,有效地去除虚警、保留目标,其耗时仅为常用方法的1/3。
In view of the problem that current methods cannot reach a good balance between capability of discrimination,utility and computational complexity,the authors have proposed in this paper an algorithm based on hierarchical feature description. Firstly,simple shape or geometrical features are extracted to get rid of large numbers of false- alarm targets based on weighted voting. Secondly,complex discrimination features are selected to form the optimal feature set by feature separation. And then the feature set is used to support vector machine to get the real ship target. Experimental results show that the proposed algorithm in this paper,which extracts hierarchical features to certain regions identified,can effectively eliminate false alarms,reduce the amount of computation,and improve accuracy and efficiency of discrimination,and can also reduce the influence of external factors,remove false alarm and reserve the targets effectively,with time spending being only 1 /3 of the common method.
In view of the problem that current methods cannot reach a good balance between capability of discrimination,utility and computational complexity,the authors have proposed in this paper an algorithm based on hierarchical feature description. Firstly,simple shape or geometrical features are extracted to get rid of large numbers of false- alarm targets based on weighted voting. Secondly,complex discrimination features are selected to form the optimal feature set by feature separation. And then the feature set is used to support vector machine to get the real ship target. Experimental results show that the proposed algorithm in this paper,which extracts hierarchical features to certain regions identified,can effectively eliminate false alarms,reduce the amount of computation,and improve accuracy and efficiency of discrimination,and can also reduce the influence of external factors,remove false alarm and reserve the targets effectively,with time spending being only 1 /3 of the common method.
引文
[1]Corbane C,Najman L,Pecoul E et al.A complete processing chain for ship detection using optical satellite imagery[J].International Journal of Remote Sensing,2010,31(22):5837-5854.
[2]Bi F K,Liu F,Gao L N.A hierarchical salient-region based algorithm for ship detection in remote sensing images[J].Lecture Notes in Electrical Engineering,2010,67:729-738.
[3]Li W W.Detection of Ship in Optical Remote Sensing Image of Median-low Resolution[D].Changsha:National University of Defense Technology,2008:19-21.
[4]Lu C Y,Zou H X,Sun H,et al.Combing rough set and RBF neural network for large-scale ship recognition in optical satellite images[C]//Proceedings of the 35th International Symposium on Remote Sensing of Environment(ISRSE35).IOP Conference Series:Earth and Environmental Science,SCI,2014,17(1).
[5]李禹,王世晞,计科峰,等.一种新的高分辨率SAR图像目标自动鉴别方法[J].国防科技大学学报,2007,29(3):81-84.Li Y,Wang S X,Ji K F,et al.A new method of automatic target discrimination in high-resolution SAR image[J].Journal of National University of Defense Technology,2007,29(3):81-84.
[6]Gonzalez R C,Woods R E,Eddins S L.Digital Image Processing Using MATLAB[M].Translated by Ruan Q Q.Beijing:Publishing House of Electronics Industry,2005:315-319.
[7]刘凯.基于分形几何理论的虹膜识别算法研究[D].济南:山东大学,2011:13-20.Liu K.The Research of Iris Recognition Algorithms Based on Fractal Geometry[D].Ji’nan:Shandong University,2011:13-20.
[8]许军毅.光学卫星遥感图像舰船目标检测技术研究[D].长沙:国防科技大学,2011:73-80.Xu J Y.The Study of Ship Target Detection in Optical Satellite Remote Sensing Image[D].Changsha:National University of Defense Technology,2011:73-80.
[9]Delphine C M.Ship detection with spaceborne multichannel SAR/GMTI radars[C]//Proceedings of 9th European Conference on Synthetic Aperture Radar.Piscataway,NJ,USA;IEEE,2012:400-403.
[10]Gao G.An improved scheme for target discrimination in high-resolution SAR images[J].IEEE Transaction on Geosciences and Remote Sensing,2011,49(1):277-294.
[11]Dardas N H,Georganas N D.Real-time hand gesture detection and recognition using bag-of-features and support vector machine techniques[J].IEEE Transactions on Instrumentation and Measurement,2011,60(11):3592-3607.
[2]Bi F K,Liu F,Gao L N.A hierarchical salient-region based algorithm for ship detection in remote sensing images[J].Lecture Notes in Electrical Engineering,2010,67:729-738.
[3]Li W W.Detection of Ship in Optical Remote Sensing Image of Median-low Resolution[D].Changsha:National University of Defense Technology,2008:19-21.
[4]Lu C Y,Zou H X,Sun H,et al.Combing rough set and RBF neural network for large-scale ship recognition in optical satellite images[C]//Proceedings of the 35th International Symposium on Remote Sensing of Environment(ISRSE35).IOP Conference Series:Earth and Environmental Science,SCI,2014,17(1).
[5]李禹,王世晞,计科峰,等.一种新的高分辨率SAR图像目标自动鉴别方法[J].国防科技大学学报,2007,29(3):81-84.Li Y,Wang S X,Ji K F,et al.A new method of automatic target discrimination in high-resolution SAR image[J].Journal of National University of Defense Technology,2007,29(3):81-84.
[6]Gonzalez R C,Woods R E,Eddins S L.Digital Image Processing Using MATLAB[M].Translated by Ruan Q Q.Beijing:Publishing House of Electronics Industry,2005:315-319.
[7]刘凯.基于分形几何理论的虹膜识别算法研究[D].济南:山东大学,2011:13-20.Liu K.The Research of Iris Recognition Algorithms Based on Fractal Geometry[D].Ji’nan:Shandong University,2011:13-20.
[8]许军毅.光学卫星遥感图像舰船目标检测技术研究[D].长沙:国防科技大学,2011:73-80.Xu J Y.The Study of Ship Target Detection in Optical Satellite Remote Sensing Image[D].Changsha:National University of Defense Technology,2011:73-80.
[9]Delphine C M.Ship detection with spaceborne multichannel SAR/GMTI radars[C]//Proceedings of 9th European Conference on Synthetic Aperture Radar.Piscataway,NJ,USA;IEEE,2012:400-403.
[10]Gao G.An improved scheme for target discrimination in high-resolution SAR images[J].IEEE Transaction on Geosciences and Remote Sensing,2011,49(1):277-294.
[11]Dardas N H,Georganas N D.Real-time hand gesture detection and recognition using bag-of-features and support vector machine techniques[J].IEEE Transactions on Instrumentation and Measurement,2011,60(11):3592-3607.