计算机科学 ›› 2019, Vol. 46 ›› Issue (9): 265-270.doi: 10.11896/j.issn.1002-137X.2019.09.040

• 图形图像与模式识别 • 上一篇    下一篇

基于多路卷积神经网络的多光谱场景识别方法

江泽涛1,2, 秦嘉奇1, 胡硕3   

  1. (桂林电子科技大学广西图像图形智能处理高校重点实验室 广西 桂林541004)1;
    (桂林电子科技大学广西可信软件重点实验室 广西 桂林541004)2;
    (南昌航空大学 南昌330063)3
  • 收稿日期:2018-08-03 出版日期:2019-09-15 发布日期:2019-09-02
  • 通讯作者: 秦嘉奇(1993-),男,硕士生,主要研究方向为计算机视觉,E-mail:1445155606@qq.com
  • 作者简介:江泽涛(1961-),男,博士,教授,CCF会员,主要研究方向为计算机视觉;胡 硕(1983-),博士生,讲师,主要研究方向为智能计算。
  • 基金资助:
    国家自然科学基金(61572147),广西科技计划项目(AC16380108),广西图像图形智能处理重点实验室项目(GIIP201701),广西研究生教育创新计划资助项目(2018YJCX46),江西省自然科学基金资助项目(20171BAB212015)

Multi-spectral Scene Recognition Method Based on Multi-way Convolution Neural Network

JIANG Ze-tao1,2, QIN Jia-qi1, HU Shuo3   

  1. (The Key Laboratory of Image and Graphic Intelligent Processing of Higher Education in Guangxi,Guilin University of Electronic Technology,Guilin,Guangxi 541004,China)1;
    (The Key Laboratory of Dependable Software of Guangxi,Guilin University of Electronic Technology,Guilin,Guangxi 541004,China)2;
    (Nanchang Hangkong University,Nanchang 330063,China)3
  • Received:2018-08-03 Online:2019-09-15 Published:2019-09-02

摘要: 现有的基于卷积神经网络的场景识别算法无法处理目标场景图形是多光谱图像的情况,在数据量较小的情况下,该算法的识别率不高。针对以上问题,提出一种基于多路卷积神经网络的多光谱场景识别方法。多路卷积神经网络接受三通道可见光彩色图像(RGB图像)以及单通道的近红外图像(NIR图像)共四通道输入。所提方法能够有效提取可见光图像特征、红外光图像特征以及可见光和红外光图像之间的关联特征,并将特征在全连接层进行融合,合理利用了各个光谱图像之间的相关信息,并通过结合预训练的方法来提高识别精度。在NIR_RGB数据集上的实验表明,与AlexNet、InceptionNet、ResNet以及人工设计特征描述子方法相比,该网络的平均识别率较高。并且,对此网络稍加改动,就能将其推广到其他多光谱图像分类任务中。

关键词: 多光谱, 卷积神经网络, 图像分类, 场景识别

Abstract: The existing scene recognition algorithm based on convolution neural network can’t deal with the multi spectral image of the target scene and can’t implement ideal accuracy in the case of insufficient data.In view of the above problems,this paper proposed a multi-spectral convolution neural network based multispectral scene recognition me-thod.The multi-way convolution neural network accepts three channels of visible light color image (RGB image) and a single channel near infrared image (NIR image) with a total of four channels.The proposed method can effectively extract the features of visible light image,infrared image and the correlation between visible and infrared images,and combine the features in the full connection layer,utilizing the correlation information among spectral images reasonably.The pre-training method is combined to improve the accuracy.Experiment results on the NIR_RGB dataset show that the average accuracy of the network is higher than that of AlexNet,InceptionNet,ResNet and artificial design feature descriptors.Moreover,this network can be extended to other multi-spectral image classification tasks with slight modification.

Key words: Multi-spectral, Convolution neural network, Image classification, Scene recognition

中图分类号: 

  • TP391.4
[1]HINTON G E,OSINDERO S,TEH Y W.A Fast Learning Algorithm For Deep Belief Nets[J].Neural Computation,2006,18(7):1527-1554.
[2]KRIZHEVSKY A,SUTSKEVER I,HINTON G E.ImageNet classification with deep convolutional neural networks[C]//International Conference on Neural Information Processing Systems.Curran Associates Inc.,2012:1097-1105.
[3]SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[C]//Proceedings of the International Conference on Learning Representations.Oxford,USA,2015.
[4]SZEGEDY C,LIU W,JIA Y,et al.Going deeper with convolutions[C]//IEEE Conference on Computer Vision and Pattern Recognition.IEEE Computer Society,2015:1-9.
[5]HE K M,ZHANG X Y,REN S Q,et al.Deep Residual Learning for Image Recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition.IEEE Computer Society,2015:770-778.
[6]GAO H,ZHUANG L,MAATEN L V D,et al.Densely Connected Convolutional Networks[C]//IEEE Conference on Computer Vision & Pattern Recognition.IEEE Computer Society,2017.
[7]HU J,SHEN L,SUN G.Squeeze-and-excitation networks[J/OL].https://arxiv.org/abs/1709.01507.
[8]ROUTRAY S,RAY A K,MISHRA C,et al.Efficient hybrid image denoising scheme based on SVM classification[J].Optik,2018,157:503-511.
[9]LEE S J,CHEN T,YU L,et al.Image Classification Based on the Boost Convolutional Neural Network[J].IEEE Access,2018,6(99):12755-12768.
[10]SHI L,WANG Y M,CAO Y J,et al.Vehicle Type Recognition Based on Deep Convolution Neural Network[J].Computer Science,2018,45(5):280-284.
[11]GONG A,FEI F,ZHENG J.Multi-person Behavior Recognition Method Based on Convolutional Neural Networks[J].Computer Science,2018,45(2):306-311,321.(in Chinese)龚安,费凡,郑君.基于卷积神经网络的多人行为识别方法[J].计算机科学,2018,45(2):306-311,321.
[12]ZHOU F Y,JIN L P,DONG J.Review of convolutional neuralnetwork[J].Chinese Journal of Computers,2017,40(6):1229-1251.(in Chinese)周飞燕,金林鹏,董军.卷积神经网络研究综述[J].计算机学报,2017,40(6):1229-1251.
[13]BROWN M,SÜSSTRUNK S.Multi-spectral SIFT for scenecategory recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition.IEEE Computer Society,2011:177-184.
[14]XIAO Y,WU J,YUAN J.mCENTRIST:A Multi-Channel Feature Generation Mechanism for Scene Categorization[J].IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society,2014,23(2):823-836.
[15]VAPNIK V N.An overview of statistical learning theory[J].IEEE Transactions on Neural Networks,1999,10(5):988-999.
[16]SALAMATI N,LARLUS D,CSURKA G.Combining Visibleand Near-Infrared Cues for Image Categorisation[C]//British Machine Vision Conference.BMVA Press,2011.
[17]ZHANG M M,CHOI J,DANIILIDIS K,et al.VAIS:A dataset for recognizing maritime imagery in the visible and infrared spectrums[C]//Computer Vision and Pattern Recognition Workshops.IEEE,2015:10-16.
[18]LIU F,SHEN T S,MA X X,et al.Ship recognition based on multi-band neural network[J].Optics and Precision Enginee-ring,2017,25(11):2939-2946.(in Chinese)刘峰,沈同圣,马新星,等.基于多波段深度神经网络的舰船目标识别[J].光学精密工程,2017,25(11):2939-2946.
[19]LIU F,SHEN T S,MA X X.Convolutional Neural NetworkBased Multi-Band Ship Target Recognition with Feature Fusion[J].Acta Optica Sinica,2017(10):240-248.(in Chinese)刘峰,沈同圣,马新星.特征融合的卷积神经网络多波段舰船目标识别[J].光学学报,2017(10):240-248.
[20]HAJHASSAN H,CHADDAD A,HARKOUSS Y,et al.Classifications of Multispectral Colorectal Cancer Tissues Using Convolution Neural Network[J].Journal of Pathology Informatics,2017,8(1):1-7.
[21]WANG J,LUO C,HUANG H,et al.Transferring Pre-Trained Deep CNNs for Remote Scene Classification with General Features Learned from Linear PCA Network[J].Remote Sensing,2017,9(3):225.
[22]LOWE D G.Distinctive image features from scale-invariant keypoint[J].International Journal of Computer Vision,2004,60(2):91-110.
[23]LEUTENEGGER S,CHLI M,SIEGWART R Y.BRISK:Binary Robust invariant scalable keypoints[C]//IEEE International Conference on Computer Vision.IEEE,2012:2548-2555.
[1] 王嫣然, 陈清亮, 吴俊君. 面向复杂环境的图像语义分割方法综述[J]. 计算机科学, 2019, 46(9): 36-46.
[2] 孙中锋, 王静. 用于基于方面情感分析的RCNN-BGRU-HN网络模型[J]. 计算机科学, 2019, 46(9): 223-228.
[3] 缪永伟, 李高怡, 鲍陈, 张旭东, 彭思龙. 基于卷积神经网络的图像局部风格迁移[J]. 计算机科学, 2019, 46(9): 259-264.
[4] 石晓红, 黄钦开, 苗佳欣, 苏卓. 基于卷积网络的边缘保持滤波方法[J]. 计算机科学, 2019, 46(9): 277-283.
[5] 于洋, 李世杰, 陈亮, 刘韵婷. 基于改进YOLO v2的船舶目标检测方法[J]. 计算机科学, 2019, 46(8): 332-336.
[6] 张琳娜,陈建强,陈晓玲,岑翼刚,阚世超. 面向行车视频目标实时检测的轻量级SSD网络[J]. 计算机科学, 2019, 46(7): 233-237.
[7] 孔繁钰, 周愉峰, 陈纲. 基于时空特征挖掘的交通流量预测方法[J]. 计算机科学, 2019, 46(7): 322-326.
[8] 陈思文, 刘玉江, 刘冬, 苏晨, 赵地, 钱林学, 张佩珩. 基于AlexNet模型和自适应对比度增强的乳腺结节超声图像分类[J]. 计算机科学, 2019, 46(6A): 146-152.
[9] 石雨鑫, 邓洪敏, 郭伟林. 基于混合卷积神经网络的静态手势识别[J]. 计算机科学, 2019, 46(6A): 165-168.
[10] 侯媛媛, 何儒汉, 李敏, 陈佳. 结合卷积神经网络多层特征融合和K-Means聚类的服装图像检索方法[J]. 计算机科学, 2019, 46(6A): 215-221.
[11] 刘泽康, 孙华志, 马春梅, 姜丽芬. 基于卷积神经网络多特征联合的车辆识别模型[J]. 计算机科学, 2019, 46(6A): 254-258.
[12] 吕培建, 陈佳鹏, 袁飞, 彭强, 项煜. 基于上下文以及多尺度信息融合的目标检测算法[J]. 计算机科学, 2019, 46(6A): 279-283.
[13] 郑诚, 洪彤彤, 薛满意. 用于短文本分类的BLSTM_MLPCNN模型[J]. 计算机科学, 2019, 46(6): 206-211.
[14] 华臻, 张海程, 李晋江. 基于残差的端对端图像超分辨率[J]. 计算机科学, 2019, 46(6): 246-255.
[15] 陈深进, 薛洋. 基于改进卷积神经网络的短时公交客流预测[J]. 计算机科学, 2019, 46(5): 175-184.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
[1] 雷丽晖,王静. 可能性测度下的LTL模型检测并行化研究[J]. 计算机科学, 2018, 45(4): 71 -75, 88 .
[2] 夏庆勋,庄毅. 一种基于局部性原理的远程验证机制[J]. 计算机科学, 2018, 45(4): 148 -151, 162 .
[3] 孙启,金燕,何琨,徐凌轩. 用于求解混合车辆路径问题的混合进化算法[J]. 计算机科学, 2018, 45(4): 76 -82 .
[4] 刘琴. 计算机取证过程中基于约束的数据质量问题研究[J]. 计算机科学, 2018, 45(4): 169 -172 .
[5] 廖星,袁景凌,陈旻骋. 一种自适应权重的并行PSO快速装箱算法[J]. 计算机科学, 2018, 45(3): 231 -234, 273 .
[6] 施超,谢在鹏,柳晗,吕鑫. 基于稳定匹配的容器部署策略的优化[J]. 计算机科学, 2018, 45(4): 131 -136 .
[7] 朱淑芹,王文宏,李俊青. 针对基于感知器模型的混沌图像加密算法的选择明文攻击[J]. 计算机科学, 2018, 45(4): 178 -181, 189 .
[8] 戴文静, 袁家斌. 隐含子群问题的研究现状[J]. 计算机科学, 2018, 45(6): 1 -8 .
[9] 邓霞, 常乐, 梁俊斌, 蒋婵. 移动机会网络组播路由的研究进展[J]. 计算机科学, 2018, 45(6): 19 -26 .
[10] 吴伟男, 刘建明. 面向低功耗无线传感器网络的动态重传算法[J]. 计算机科学, 2018, 45(6): 96 -99,123 .