兼顾特征级和决策级融合的场景分类

doi:10.11772/j.issn.1001-9081.2016.05.1262

计算机应用 ›› 2016, Vol. 36 ›› Issue (5): 1262-1266.DOI: 10.11772/j.issn.1001-9081.2016.05.1262

兼顾特征级和决策级融合的场景分类

何刚^1,2, 霍宏^1,2, 方涛^1,2

1. 上海交通大学自动化系, 上海 200240;
2. 系统控制与信息处理教育部重点实验室, 上海 200240

收稿日期:2015-10-21 修回日期:2015-12-14 出版日期:2016-05-10 发布日期:2016-05-09
通讯作者: 霍宏
作者简介:何刚(1991-),男,四川南充人,硕士研究生,主要研究方向:模式识别、场景分类;霍宏(1972-),女,辽宁本溪人,讲师,博士,主要研究方向:模式识别、遥感图像理解;方涛(1965-),男,四川彭山人,教授,博士,主要研究方向:图像理解、遥感科学。
基金资助:
国家973计划项目(2012CB719903);国家自然科学基金委创新研究群体资助项目(61221003);国家自然科学基金青年科学基金资助项目(41101386)。

Scene classification based on feature-level and decision-level fusion

HE Gang^1,2, HUO Hong^1,2, FANG Tao^1,2

1. Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China;
2. Key Laboratory of System Control and Information Processing, Ministry of Education, Shanghai 200240, China

Received:2015-10-21 Revised:2015-12-14 Online:2016-05-10 Published:2016-05-09
Supported by:
This work is partially supported by the National Basic Research Program (973 Program) of China (2012CB719903), the Science Fund for Creative Research Groups of the National Natural Science Foundation of China (61221003),and the Young Scientists Fund of the National Natural Science Foundation of China (41101386).

摘要/Abstract

摘要： 针对单一特征在场景分类中精度不高的问题,借鉴信息融合的思想,提出了一种兼顾特征级融合和决策级融合的分类方法。首先,提取图像的尺度不变特征变换词包(SIFT-BoW)、Gist、局部二值模式(LBP)、Laws纹理以及颜色直方图五种特征。然后,将每种特征单独对场景进行分类得到的结果以Dezert-Smarandache理论(DSmT)推理的方式在决策级进行融合,获得决策级融合下的分类结果;同时,将五种特征串行连接实现特征级融合并进行分类,得到特征级融合下的分类结果。最后,将特征级和决策级的分类结果进行自适应的再次融合完成场景分类。在决策级融合中,为解决DSmT推理过程中基本信度赋值(BBA)构造困难的问题,提出一种利用训练样本构造后验概率矩阵来完成基本信度赋值的方法。在21类遥感数据集上进行分类实验,当训练样本和测试样本各为50幅时,分类精度达到88.61%,较单一特征中的最高精度提升了12.27个百分点,同时也高于单独进行串行连接的特征级融合或DSmT推理的决策级融合的分类精度。

关键词: 场景分类, 特征级融合, 决策级融合, Dezert-Smarandache理论推理, 基本信度赋值, 遥感影像

Abstract: Since the accuracy of single feature in scene classification is low, inspired by information fusion, a classification method combined feature-level and decision-level fusion was proposed. Firstly, Scale Invariant Feature Transform-Bag of Words (SIFT-BoW), Gist, Local Binary Patterns (LBP), Laws texture and color histogram features of image were extracted. Then, the classification results of every single feature were fused in the way of Dezert-Smarandache Theory (DSmT) to obtain the decision-level fusion result; at the same time, the five features were serially connected to generate a new feature, the new feature was used to classification to obtain the feature-level fusion result. Finally, the feature-level and decision-level fusion results were adaptively fused to finish classification. To solve the Basic Belief Assignment (BBA) problem of DSmT, a method based on posterior probability matrix was proposed. The accuracy of the proposed method on 21 classes of remote sensing images is 88.61% when training and testing samples are both 50, which is 12.27 percentage points higher than the highest accuracy of single feature. The accuracy of proposed method is also higher than that of the feature-level fusion serial connection or DSmT reasoning decision-level fusion.

Key words: scene classification, feature-level fusion, decision-level fusion, Dezert-Smarandache Theory (DSmT) reasoning, Basic Belief Assignment (BBA), remote sensing image

中图分类号:

TP391.4

何刚, 霍宏, 方涛. 兼顾特征级和决策级融合的场景分类[J]. 计算机应用, 2016, 36(5): 1262-1266.

HE Gang, HUO Hong, FANG Tao. Scene classification based on feature-level and decision-level fusion[J]. Journal of Computer Applications, 2016, 36(5): 1262-1266.

参考文献

[1] CHERIVADAT A M. Unsupervised feature learning for aerial scene classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2014, 52(1):439-451.
[2] 李学龙, 史建华, 董永生, 等. 场景图像分类技术综述[J].中国科学:信息科学, 2015, 45(7):827-848. (LI X L, SHI J H, DONG Y S, et al. A survey on scene image classification[J]. Scientia Sinica:Informations, 2015, 45(7):827-848.)
[3] ZHOU L, ZHOU Z, HU D. Scene classification using multi-resolution low-level feature combination[J]. Neurocomputing, 2013, 122:284-297.
[4] KAPELA R, MCGUINNESS K, O'CONNOR N E. Real-time field sports scene classification using colour and frequency space decompositions[J/OL].(2014-06-08)[2015-10-21].DOI 10.1007/s11554-014-0437-7.
[5] 李锦锋, 许勇. 基于LBP和小波纹理特征的室内室外场景分类算法[J]. 中国图象图形学报, 2010, 15(5):742-748.(LI J F, XU Y. Indoor-outdoor scene classification algorithm based on the texture feature of LBP and wavelet[J]. Journal of Image and Graphics, 2010, 15(5):742-748.)
[6] OLIVA A, TORRALBA A. Modeling the shape of the scene:a holistic representation of the spatial envelope[J]. International Journal of Computer Vision, 2001, 42(3):145-175.
[7] SIVIC J, ZISSERMAN A. Video Google:a text retrieval approach to object matching in videos[C]//Proceedings of the 9th IEEE International Conference on Computer Vision. Washington, DC:IEEE Computer Society, 2003:1470-1478.
[8] BOLOYINOU A, PRATIKAKIS I, PERANTONIS S. Bag of spatio-visual words for context inference in scene classification[J]. Pattern Recognition, 2013, 46(3):1039-1053.
[9] SHAO W, YANGW, XIA G S, et al. A hierarchical scheme of multiple feature fusion for high-resolution satellite scene categorization[C]//Proceedings of the 9th International Conference on Computer Vision Systems. Berlin:Springer-Verlag, 2013:324-333.
[10] YU J, TAO D, RUI Y, et al. Pairwise constraints based multiview features fusion for scene classification[J]. Pattern Recognition, 2013, 46(2):483-496.
[11] 肖保良. 基于Gist特征与PHOG特征融合的多类场景分类[J]. 中北大学学报(自然科学版), 2014, 35(6):690-694.(XIAO B L. Multiclass scene classification based on fusion of Gist feature and PHOG feature[J]. Journal of North University of China (Natural Science Edition), 2014, 35(6):690-694.)
[12] 刘帅, 李士近, 冯钧. 多特征融合的遥感图像分类[J]. 数据采集与处理, 2014, 29(1):108-115.(LIU S, LI S J, FENG J. Remote sensing image classification based on adaptive fusion of multiple features[J]. Journal of Data Acquisition and Processing, 2014, 29(1):108-115.)
[13] OJALA T, PIETIKAINEN M, HARWOOD D. A comparative study of texture measures with classification based on feature distributions[J]. Pattern Recognition, 1996, 29(1):51-59.
[14] OJALA T, PIETIKAINEN M, MAENPAA T. Multiresolution gray-scale and rotation invariant texture classification with local binary patterns[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, 24(7):971-987.
[15] ZHAO G, AHONEN T, MATAS J, et al. Rotation-invariant image and video description with local binary pattern features[J].IEEE Transactions on Image Processing, 2012, 21(4):1465-1477.
[16] SETIAWAN A S, WESLEY J, PURNAMA Y. Mammogram classification using Law's texture energy measure and neural networks[J]. Procedia Computer Science, 2015, 59:92-97.
[17] KHAN R, HANBURY A, STOTTINGER J, et al. Multiple color space channel fusion for skin detection[J]. Multimedia Tools and Applications, 2014, 72(2):1709-1730.
[18] 李强. 手部特征识别及特征级融合算法研究[D]. 北京:北京交通大学, 2006:2-15.(LI Q. Research on handmetric recognition and feature level fusion method[D]. Beijing:Beijing Jiaotong University, 2006:2-15.)
[19] DEZERT J. Foundations for a new theory of plausible and paradoxical reasoning[J]. Information & Security, 2002, 9(1):13-57.
[20] SMARANDACHE F, DEZERT J. Advances and Applications of DSmT for Information Fusion[M]. Rehoboth:American Research Press, 2006:36-50.
[21] DEZERT J, LIU Z, MERCIER G. Edge detection in color images based on DSmT[C]//Proceedings of the 14th IEEE International Conference on Information Fusion. Piscataway, NJ:IEEE, 2011:1-8.
[22] TCHAMOVA A, DEZERT J. Intelligent alarm classification based on DSmT[C]//Proceedings of the 6th IEEE International Conference on Intelligent Systems. Piscataway, NJ:IEEE, 2012:120-125.

兼顾特征级和决策级融合的场景分类

Scene classification based on feature-level and decision-level fusion

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	宋中山, 梁家锐, 郑禄, 刘振宇, 帖军. 基于双向门控尺度特征融合的遥感场景分类[J]. 计算机应用, 2021, 41(9): 2726-2735.
[2]	任秋如, 杨文忠, 汪传建, 魏文钰, 钱芸芸. 遥感影像变化检测综述[J]. 计算机应用, 2021, 41(8): 2294-2305.
[3]	任媛媛, 汪传建. 数据异常情况下遥感影像时间序列分类算法[J]. 计算机应用, 2021, 41(3): 662-668.
[4]	王睿, 黄微, 胡南强. 基于全变分模型的多时相遥感影像厚云去除算法[J]. 计算机应用, 2020, 40(7): 2126-2130.
[5]	边小勇, 费雄君, 穆楠. 基于尺度注意力网络的遥感图像场景分类[J]. 计算机应用, 2020, 40(3): 872-877.
[6]	王玉龙, 蒲军, 赵江华, 黎建辉. 基于生成对抗网络的地面新增建筑检测[J]. 计算机应用, 2019, 39(5): 1518-1522.
[7]	冯国政, 徐金东, 范宝德, 赵甜雨, 朱萌, 孙潇. 基于半监督模糊C均值算法的遥感影像分类[J]. 计算机应用, 2019, 39(11): 3227-3232.
[8]	汪月云, 黄微, 王睿. 基于薄云厚度分布评估的遥感影像高保真薄云去除方法[J]. 计算机应用, 2018, 38(12): 3596-3600.
[9]	李金蔓, 汪剑鸣, 金光浩. 基于特征级和决策级融合的人脸吸引力评价方法[J]. 计算机应用, 2018, 38(12): 3607-3611.
[10]	叶利华, 王磊, 赵利平. 低小慢无人机降落野外场景识别方法[J]. 计算机应用, 2017, 37(7): 2008-2013.
[11]	施文灶, 刘金清. 基于邻域总变分和势直方图函数的高分辨率遥感影像建筑物提取[J]. 计算机应用, 2017, 37(6): 1787-1792.
[12]	严考碧, 李志欣, 张灿龙. 基于主题模型的多示例多标记学习方法[J]. 计算机应用, 2015, 35(8): 2233-2237.
[13]	顾秀颖, 赵子沂, 方涛, 霍宏. 高分辨率遥感影像复杂场景的汇总统计方法[J]. 计算机应用, 2015, 35(3): 849-853.
[14]	陆惠玲, 周涛, 王惠群, 王文文. 基于特征级融合神经网络的磁共振成像前列腺肿瘤CAD模型[J]. 计算机应用, 2015, 35(10): 2813-2818.
[15]	李博曹鹏栗伟赵大哲. 基于尺度空间中多特征融合的医学影像分类[J]. 计算机应用, 2013, 33(04): 1108-1111.