正在加载图片...
第11卷第5期 智能系统学报 Vol.11 No.5 2016年10月 CAAI Transactions on Intelligent Systems 0ct.2016 D0I:10.11992/is.201511011 网络出版地址:htp:/nww.cnki.net/kcms/detail/23.1538.TP.20160824.0929.012.html 多特征的光学遥感图像多目标识别算法 姬晓飞,秦宁丽2,刘洋1 (1.沈阳航空航天大学自动化学院,辽宁沈阳110136:2.北京国电通网络技术有限公司,北京100070) 摘要:基于单一特征的光学遥感图像多目标分类识别存在准确性较差的问题,提出一种新的基于多特征决策级融 合的多目标分类识别算法。首先对光学遥感图像目标提取3种能够同时满足平移、旋转和尺度不变性的特征:可以 描述局部和全局分布特性的分层BoF-SIFT特征,描述目标边缘轮廓点信息的改进后的SC形状特征,对图像中较大 目标识别较好的H山不变矩特征:其次采用基于径向基核函数的一对一支持向量机算法分别获得3种特征的目标识 别概率,并设计了一种多特征决策级加权融合的策略实现对多目标的分类。经多次实验验证该算法对光学遥感图 像多目标具有较好的分类识别性能,达到了93.52%的正确识别率。 关键词:光学遥感图像;多特征的决策级融合;分层的BoF-SIFT特征:SC形状特征;Hu不变矩特征;支持向量机 中图分类号:TP751.1文献标志码:A文章编号:1673-4785(2016)05-0655-08 中文引用格式:姬晓飞,秦宁丽,刘洋.多特征的光学遥感图像多目标识别算法研究[J].智能系统学报,2016,11(5):655-662. 英文引用格式:JI Xiaofei,QIN Ningli,LIU Yang.Research on multi-feature based multi--target recognition algorithm for optical re- mote sensing image[J].CAAI transactions on intelligent systems,2016,11(5):655-662. Research on multi-feature based multi-target recognition algorithm for optical remote sensing image JI Xiaofei',QIN Ningli2,LIU Yang' (1.School of Automation,Shenyang Aerospace University,Shenyang 110136,China;2.Beijing GuoDianTong Network Technology Co. Ltd,Beijing 100070,China) Abstract:A novel multi-feature decision level fusion recognition algorithm is proposed to solve the problem of poor levels of accuracy with single feature based multi-target classification of optical remote sensing images.Firstly,three kinds of features which can not only meet translation,rotation,and scale invariance are extracted.One is the hier- archical BoF-SIFT feature which can simultaneously describe local and global distributions.Another is the improved shape context feature which is used to describe the target edge contour point information.The other one is Hu mo- ment invariants which gives better levels of recognition performance for large targets.Secondly,the recognition probabilities of these features are obtained using a one versus one support vector machine based on a radial basis function.Thirdly a strategy for multi-feature decision level fusion is designed.A large number of experiments show that the algorithm for multi-target classification of optical remote sensing images performs better with the recognition rate of targets reaching 93.52%. Keywords:optical remote sensing image;multi-features decision level fusion;hierarchical BoF-SIFT feature; shape context feature;Hu moment invariants;support vector machine 随着遥感技术和模式识别技术的不断发展,遥 标分类识别,已经成为了遥感图像处理和分析领域 感图像目标分类识别,尤其是光学遥感图像中的目 备受关注的重要方向山。 光学遥感图像目标识别算法的准确性很大程度 收稿日期:2015-11-10.网络出版日期:2016-08-24. 基金项目:国家自然科学基金项目(61103123):辽宁省高等学校优秀人 上取决于所提取特征的适应性。文献[2]提出了基 才支持计划项目(LJQ214018):辽宁省自然科学基金项目 (2015020101). 于自组织特征映射(self-organizing feature map, 通信作者:姬晓飞.E-mail:jixiaofei7804@126.com. SOFM)网络模型的纹理分类算法,该算法仅对飞机第 11 卷第 5 期 智 能 系 统 学 报 Vol.11 №.5 2016 年 10 月 CAAI Transactions on Intelligent Systems Oct. 2016 DOI:10.11992 / tis.201511011 网络出版地址:http: / / www.cnki.net / kcms/ detail / 23.1538.TP.20160824.0929.012.html 多特征的光学遥感图像多目标识别算法 姬晓飞1 ,秦宁丽2 ,刘洋1 (1.沈阳航空航天大学 自动化学院,辽宁 沈阳 110136; 2.北京国电通网络技术有限公司,北京 100070) 摘 要:基于单一特征的光学遥感图像多目标分类识别存在准确性较差的问题,提出一种新的基于多特征决策级融 合的多目标分类识别算法。 首先对光学遥感图像目标提取 3 种能够同时满足平移、旋转和尺度不变性的特征:可以 描述局部和全局分布特性的分层 BoF⁃SIFT 特征,描述目标边缘轮廓点信息的改进后的 SC 形状特征,对图像中较大 目标识别较好的 Hu 不变矩特征;其次采用基于径向基核函数的一对一支持向量机算法分别获得 3 种特征的目标识 别概率,并设计了一种多特征决策级加权融合的策略实现对多目标的分类。 经多次实验验证该算法对光学遥感图 像多目标具有较好的分类识别性能,达到了 93.52%的正确识别率。 关键词:光学遥感图像;多特征的决策级融合;分层的 BoF⁃SIFT 特征;SC 形状特征;Hu 不变矩特征;支持向量机 中图分类号:TP751.1 文献标志码:A 文章编号:1673⁃4785(2016)05⁃0655⁃08 中文引用格式:姬晓飞,秦宁丽,刘洋.多特征的光学遥感图像多目标识别算法研究[J]. 智能系统学报, 2016, 11(5): 655⁃662. 英文引用格式:JI Xiaofei, QIN Ningli, LIU Yang.Research on multi⁃feature based multi⁃target recognition algorithm for optical re⁃ mote sensing image[J]. CAAI transactions on intelligent systems, 2016,11(5):655⁃662. Research on multi⁃feature based multi⁃target recognition algorithm for optical remote sensing image JI Xiaofei 1 , QIN Ningli 2 , LIU Yang 1 (1.School of Automation, Shenyang Aerospace University, Shenyang 110136, China; 2.Beijing GuoDianTong Network Technology Co. Ltd, Beijing 100070, China) Abstract:A novel multi⁃feature decision level fusion recognition algorithm is proposed to solve the problem of poor levels of accuracy with single feature based multi⁃target classification of optical remote sensing images. Firstly, three kinds of features which can not only meet translation, rotation, and scale invariance are extracted. One is the hier⁃ archical BoF⁃SIFT feature which can simultaneously describe local and global distributions. Another is the improved shape context feature which is used to describe the target edge contour point information. The other one is Hu mo⁃ ment invariants which gives better levels of recognition performance for large targets. Secondly, the recognition probabilities of these features are obtained using a one versus one support vector machine based on a radial basis function. Thirdly a strategy for multi⁃feature decision level fusion is designed. A large number of experiments show that the algorithm for multi⁃target classification of optical remote sensing images performs better with the recognition rate of targets reaching 93.52%. Keywords:optical remote sensing image; multi⁃features decision level fusion; hierarchical BoF⁃SIFT feature; shape context feature; Hu moment invariants; support vector machine 收稿日期:2015⁃11⁃10. 网络出版日期:2016⁃08⁃24. 基金项目:国家自然科学基金项目(61103123);辽宁省高等学校优秀人 才支持计划项目( LJQ214018);辽宁省自然科学基金项目 (2015020101). 通信作者:姬晓飞. E⁃mail:jixiaofei7804@ 126.com. 随着遥感技术和模式识别技术的不断发展,遥 感图像目标分类识别,尤其是光学遥感图像中的目 标分类识别,已经成为了遥感图像处理和分析领域 备受关注的重要方向[1] 。 光学遥感图像目标识别算法的准确性很大程度 上取决于所提取特征的适应性。 文献[2]提出了基 于自 组 织 特 征 映 射 ( self⁃organizing feature map, SOFM)网络模型的纹理分类算法,该算法仅对飞机
向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有