基于谱回归特征降维与后向传播神经网络的识别方法研究

邬战军; 牛敏; 许冰; 牛燕雄; 耿天琪; 张帆; 满达

doi:10.11999/JEIT150781

基于谱回归特征降维与后向传播神经网络的识别方法研究

doi: 10.11999/JEIT150781

计量
- 文章访问数: 710
- HTML全文浏览量: 50
- PDF下载量: 644
- 被引次数: 0
出版历程
- 收稿日期: 2015-06-29
- 修回日期: 2015-11-09
- 刊出日期: 2016-04-19

Research on Recognition Method Based on Spectral Regression and Back Propagation Neural Network

摘要

摘要: 采用后向传播(BP)神经网络对空间目标进行识别时，高维的输入特征导致网络结构复杂，识别性能降低。针对上述难点，该文提出一种基于谱回归(SR)特征降维与BP神经网络的识别方法。该方法首先对空间目标进行HOG特征提取，然后将提取的高维HOG特征进行SR降维，最后把降维后的数据通过BP分类器进行训练识别。实验结果表明：该方法的降维和识别特性优于传统降维方法PCA, KPAC, LPP, KLPP等，能够兼顾实时性和准确性，提高了识别性能。
- 目标识别 /
- 后向传播神经网络 /
- 谱回归 /
- 特征降维
Abstract: When using Back Propagation (BP) neural network to recognize the spatial target, the high dimensional input features induce the complexity of the network structure and the poor performance of the recognition. In this paper, a new recognition method based on Spectral Regression (SR) feature dimension reduction and BP neural network is proposed for the above difficulties. Firstly, the HOG features are extracted from the spatial object, and then the feature dimensions are reduced by SR. Finally, the BP classifier is used to train the data. Experimental results show that the proposed method is better than the traditional dimension reduction methods such as PCA, KPCA, LPP, KLPP in dimension reduction and recognition, which can juggle real-time and accuracy, thus improving the recognition performance.
- Target recognition /
- Back Propagation (BP) neural network /
- Spectral Regression (SR) /
- Feature dimension reduction

HTML全文

参考文献(32)

[1]	乔俊飞, 李淼, 刘江. 一种神经网络快速修剪算法[J]. 电子学报, 2010, 38(4): 830-834.
[2]	QIAO Junfei, LI Miao, and LIU Jiang. A fast pruning algorithm for neural network[J]. Acta Electronica Sinca, 2010, 38(4): 830-834.
[3]	曾岳, 冯大政. 一种基于加权变形的2DPCA的人脸特征提取方法[J]. 电子与信息学报, 2011, 33(4): 769-774. doi: 10.3724/ SP.J.1146.2010.01003.
[4]	ZENG Yue and FENG Dazheng. An algorithm of feature extraction of face based on the weighted variation of 2DPC[J]. Journal of Electronics Information Technology, 2011, 33(4): 769-774. doi: 10.3724/SP.J.1146.2010.01003.
[5]	马原, 吕群波, 刘扬阳. 基于主成分变换的图像稀疏度估计方法[J]. 物理学报, 2013, 62(20): 204202-1-204202-11. doi: 10.7498/aps.62.204202.
[6]	MA Yuan, L Qunbo, and LIU Yangyang. Image sparsity evaluation based on principle component analysis[J]. Acta Physica Sinica, 2013, 62(20): 204202-1-204202-11. doi: 10.7498/aps.62.204202.
[7]	孙韶媛, 李琳娜, 赵海涛. 采用KPCA和BP神经网络的单目车载红外图像深度估计[J]. 红外与激光工程, 2013, 42(9): 2348-2352.
[8]	SUN Shaoyuan, LI Linna, and ZHAO Haitao. Depth estimation from monocular vehicle infrared images based on KPCA and BP neural network[J]. Infrared and Laser Engineering, 2013, 42(9): 2348-2352.
[9]	ZHENG W, LAI J, and YUEN P C. Penalized preimage learning in kernel principal component analysis[J]. IEEE Transactions on Neural Networks, 2010, 21(4): 551-570.
[10]	ZHANG L M, QIAO L S, and CHEN S C. Graph-optimized locality preserving projections[J]. Pattern Recognition, 2010, 43(6): 1993-2002.
[11]	张志伟, 杨帆, 夏克文, 等. 一种有监督的LPP算法及其在人脸识别中的应用[J]. 电子与信息学报, 2008, 30(3): 539-541.
[12]	ZHANG Zhiwei, YANG Fan, XIA Kewen, et al. A supervised LPP algorithm and its application to face recognition[J]. Journal of Electronics Information Technology, 2008, 30(3): 539-541.
[13]	于攀, 叶俊勇. 基于谱回归和核空间最近邻的基因表达数据分类[J]. 电子学报, 2011, 39(8): 1955-1960.
[14]	YU Pan and YE Junyong. Spectral regression and kernel space K-nearest neighbor for classification of gene expression data[J]. Acta Electronica Sinica, 2011, 39(8): 1955-1960.
[15]	NARNEET D and BILL T. Histograms of oriented gradients for human detection[C]. Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, San Diego, 2005: 886-893.
[16]	SU X, LIN W, ZHENG X, et al. A new local-main gradient orientation HOG and contour differences based algorithm for object classification[C]. IEEE International Symposium on Circuits and Systems, Beijing, 2013: 2892-2895.
[17]	罗会兰, 钟宝康, 孔繁胜. 带权分块压缩感知的预测目标跟踪算法[J]. 电子与信息学报, 2015, 37(5): 1160-1166. doi: 10.11999/JEIT140997.
[18]	LUO Huilan, ZHONG Baokang, and KONG Fansheng. Tracking using weighted block compressed sensing and location prediction[J]. Journal of Electronics Information Technology, 2015, 37(5): 1160-1166. doi: 10.11999/ JEIT140997.
[19]	韩贵金, 朱虹. 基于HOG和颜色特征融合的人体姿态估计[J]. 模式识别与人工智能, 2014, 27(9): 769-777.
[20]	HAN Guijin and ZHU Hong. Human pose estimation based on fusion of HOG and color feature[J]. Pattern Recognition and Artificial Intelligence, 2014, 27(9): 769-777.
[21]	宋丹, 唐林波, 赵保军. 基于仿射梯度方向直方图特征的目标识别算法[J]. 电子与信息学报, 2013, 35(6): 1428-1434. doi: 10.3724/SP.J.1146.2012.01241.
[22]	SONG Dan, TANG Linbo, and ZHAO Baojun. The object recognition algorithm based on affine histogram of oriented gradient[J]. Journal of Electronics Information Technology, 2013, 35(6): 1428-1434. doi: 10.3724/SP.J.1146.2012.01241.
[23]	姜伟, 张晶, 扬炳儒. 自适应正则化核二维判别分析[J]. 模式识别与人工智能, 2014, 27(12): 1089-1097.
[24]	JIANG Wei, ZHANG Jing, and YANG Bingru. Adaptive regularization based kernel two dimensional discriminant analysis[J]. Pattern Recognition and Artificial Intelligence, 2014, 27(12): 1089-1097.
[25]	孙丽娟. 基于谱回归的人脸识别的研究[D]. [硕士论文], 重庆大学, 2009.
[26]	SUN Lijuan. Research on the face recognition based on spectral regression[D]. [Master dissertation], Chongqing University, 2009.
[27]	潘炜深, 金连文, 冯子勇. 基于多尺度梯度及深度神经网络的汉字识别[J]. 北京航空航天大学学报, 2015, 41(4): 751-756. doi: 10.13700/j.bh. 1001-5965.2014.0499.
[28]	PAN W S, JIN L W, and FENG Z Y. Recognition of Chinese characters based on multi-scale gradient and deep neural network[J]. Journal of Beijing University of Aeronautics and Astronautics, 2015, 41(4): 751-756. doi: 10.13700/j.bh. 1001-5965.2014.0499.
[29]	张海燕, 李欣, 田书峰. 基于BP神经网络的仿真线设计及其FPGA实现[J]. 电子与信息学报, 2007, 29(5): 1262-1265.
[30]	ZHANG Haiyan, LI Xin, and TIAN Shufeng. Simulation line design and its FPGA realization based on BP neural network[J]. Journal of Electronics Information Technology, 2007, 29(5): 1262-1265.
[31]	孙琼, 王光飞. 基于BP神经网络与HMM的驾驶状态识别[C]. 2014中国汽车工程学会年会, 上海, 2014: 430-432.
[32]	SUN Qiong and WANG Guangfei. Driving state recognition based on BP neural network and HMM[C]. 2014 SAE-China Congress Exhibition, Shanghai, 2014: 430-432.

相关文章(20)

[1]	李璐, 杜兰, 何浩男, 李晨, 邓盛. 基于深度森林的多级特征融合SAR目标识别, 电子与信息学报. doi: 10.11999/JEIT200685
[2]	宋晨, 周良将, 吴一戎, 丁赤飚. 基于时频集中度指标的多旋翼无人机微动特征参数估计方法, 电子与信息学报. doi: 10.11999/JEIT190309
[3]	付哲泉, 李尚生, 李相平, 但波, 王旭坤. 基于高效可扩展改进残差结构神经网络的舰船目标识别技术, 电子与信息学报. doi: 10.11999/JEIT190913
[4]	张东波, 陈治强, 易良玲, 许海霞. 图像微观结构的二值化表示与目标识别应用, 电子与信息学报. doi: 10.11999/JEIT170513
[5]	徐少坤, 刘记红, 袁翔宇. 基于HRRP序列的中段目标二维几何特征反演方法, 电子与信息学报. doi: 10.11999/JEIT161303
[6]	吴泽民, 刘涛, 姜青竹, 胡磊. 视频编码参数对目标识别性能影响的研究, 电子与信息学报. doi: 10.11999/JEIT141613
[7]	丁军, 刘宏伟, 王英华, 王正珏, 齐会娇, 时荔蕙. 一种联合阴影和目标区域图像的SAR目标识别方法, 电子与信息学报. doi: 10.11999/JEIT140713
[8]	徐少坤, 刘记红, 袁翔宇, 陆静. 基于ISAR图像的中段目标二维几何特征反演方法, 电子与信息学报. doi: 10.11999/JEIT140338
[9]	韩勋, 杜兰, 刘宏伟. 基于窄带雷达组网的空间锥体目标特征提取方法, 电子与信息学报. doi: 10.3724/SP.J.1146.2013.02037
[10]	张国亮, 杨春玲, 王暕来. 基于优化概率神经网络和红外多光谱融合的大气层外空间弹道目标识别, 电子与信息学报. doi: 10.3724/SP.J.1146.2013.00623
[11]	储岳中, 徐波, 高有涛, 邰伟鹏. 基于近邻传播聚类与核匹配追踪的遥感图像目标识别方法, 电子与信息学报. doi: 10.3724/SP.J.1146.2014.00422
[12]	李开明, 张群, 梁必帅, 罗迎, 杨小优. 卡车目标遮挡效应建模及微多普勒特征分析, 电子与信息学报. doi: 10.3724/SP.J.1146.2012.01609
[13]	王芳, 盛卫星, 马晓峰, 王昊. 基于Gabor幅值特征和相位特征相融合的ISAR像目标识别, 电子与信息学报. doi: 10.3724/SP.J.1146.2012.01500
[14]	罗迎, 柏又青, 张群, 段艳丽, 朱丰. 弹道目标平动补偿与微多普勒特征提取方法, 电子与信息学报. doi: 10.3724/SP.J.1146.2011.00607
[15]	陈琳, 卢湖川. 基于ML-pLSA模型的目标识别算法, 电子与信息学报. doi: 10.3724/SP.J.1146.2011.00455
[16]	罗迎, 张群, 封同安, 李松, 梁贤姣. OFD-LFM MIMO雷达中旋转目标微多普勒效应分析及三维微动特征提取, 电子与信息学报. doi: 10.3724/SP.J.1146.2010.00234
[17]	罗磊, 李跃华. 基于非相关判别邻域保持投影的毫米波雷达目标识别, 电子与信息学报. doi: 10.3724/SP.J.1146.2009.01534
[18]	王璐, 刘宏伟. 基于时频图的微动目标运动参数提取和特征识别的方法, 电子与信息学报. doi: 10.3724/SP.J.1146.2009.01127
[19]	王涛, 王雪松, 刘进, 肖顺平. 干扰背景下瞬时极化测量雷达的目标识别预处理技术, 电子与信息学报. doi: 10.3724/SP.J.1146.2007.01525
[20]	承德保. 基于多特征组的遥感图像中建筑物目标自动识别与标绘的方法, 电子与信息学报. doi: 10.3724/SP.J.1146.2008.00414