基于卷积神经网络与多尺度空间编码的场景识别方法

doi:10.16180/j.cnki.issn1007-7820.2020.12.011

电子科技 ›› 2020, Vol. 33 ›› Issue (12): 54-58.doi: 10.16180/j.cnki.issn1007-7820.2020.12.011

基于卷积神经网络与多尺度空间编码的场景识别方法

缪冉,李菲菲,陈虬

上海理工大学光电信息与计算机工程学院,上海 200093

收稿日期:2019-09-14 出版日期:2020-12-15 发布日期:2020-12-22
作者简介:缪冉(1995-),男,硕士研究生。研究方向:图像处理与模式识别。|李菲菲(1970-),女,博士,教授。研究方向:多媒体信息处理、图像处理与模式识别、信息检索等。|陈虬(1972-),男,博士,教授,博士生导师。研究方向:图像处理与模式识别、计算机视觉、信息检索等。
基金资助:
上海市高校特聘教授(东方学者)岗位计划(ES2015XX)

Scene Recognition Algorithm Based on Convolutional Neural Networks and Multi-Scale space Encoding

MIAO Ran,LI Feifei,CHEN Qiu

School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 20093,China

Received:2019-09-14 Online:2020-12-15 Published:2020-12-22
Supported by:
The Program for Professor of Special Appointment (Eastern Scholar) at Shanghai Institutions of Higher Learning(ES2015XX)

摘要/Abstract

摘要：

场景图像往往是由一些前景物体与背景环境以一定的空间布局组成。同类场景图片由于采样时的尺度、视角以及背景的不同而具有严重的类内差异性;存在于异类场景间的共有物体也导致异类场景图像间具有一定的相似性。据此,文中提出了基于CNN与多尺度空间编码的场景描述及识别方法。该方法结合了多尺度密集采样方法、卷积网络算法与多尺度空间编码方法。多尺度空间的编码方法是将采样网络进行多次空间划分,且对不同子区域中的CNN特征进行聚合,生成多尺度空间VLAD。文中在Scene15场景数据集上进行了实验,结果显示测试精度达到了94.67%。

关键词: 场景识别, 卷积神经网络, K均值聚类, VLAD, 主成分分析法, 支持向量机

Abstract:

A scene image is generally composed of some foreground objects and background contexts with a certain spatial layout. Due to different scales, viewpoints, and backgrounds, there exists large intra-class variation within the same scene class. On the other hand, the common objects also result in a certain inter-class similarities among heterogeneous scenes as well. Consequently, the multi-scale space encoding based on convolutional neural networks (CNN) for scene representation is proposed in the study, which combines multi-scale dense sampling method, CNN algorithm, and multi-scale space encoding method. The multi-scale encoding method spatially partitions the sampling grid many times, and then aggregates the CNN features within sub-regions with different shapes for generating the multi-scale space VLAD. The experiment is carried out on the Scene15 scene dataset, and the test results show that the test accuracy reaches 94.67%.

Key words: scene recognition, convolutional neural networks, K-means clustering, VLAD, PCA, SVM

中图分类号:

TP391

缪冉,李菲菲,陈虬. 基于卷积神经网络与多尺度空间编码的场景识别方法[J]. 电子科技, 2020, 33(12): 54-58.

MIAO Ran,LI Feifei,CHEN Qiu. Scene Recognition Algorithm Based on Convolutional Neural Networks and Multi-Scale space Encoding[J]. Electronic Science and Technology, 2020, 33(12): 54-58.

图/表 6

图1

图2

图3

图4

图5

表1

参考文献 23

[1]	Lowe D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision, 2004,60(2):91-110.
[2]	Dalal N, Triggs B. Histograms of origented gradients for human detection[C]. San Diego:Computer Vision and Pattern Recognition, 2005.
[3]	Bay H, Ess A, Tuytelaars T, et al. Speeded-Up Robust Features (SURF)[J]. Computer Vision and Image Understanding, 2008,110(3):346-359.
[4]	Sivic J, Zisserman A. Video Google:A text retrieval approach to object matching in videos[C]. Nice:International Conference on Computer Vision, 2003.
[5]	Sánchez J, Perronnin F, Mensink T, et al. Image classification with the fisher vector: theory and practice[J]. International Journal of Computer Vision, 2013,105(3):222-245.
[6]	Jégou H, Douze M, Schmid C, et al. Aggregating local descriptors into a compact image representation[C]. San Francisco:Computer Vision and Pattern Recognition, 2010.
[7]	Lazebnik S, Schmid C, Ponce J. Beyond bags of features: spatial pyramid matching for recognizing natural scene categories[C]. New York:Computer Vision and Pattern Recognition, 2006.
[8]	Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition[C]. San Diego:International Conference on Learning Representations, 2015.
[9]	Szegedy C, Liu Wei, Jia Yangqing, et al. Going deeper with convolutions[C]. Boston:Computer Vision and Pattern Recognition, 2015.
[10]	He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image recognition[C]. Las Vegas:Computer Vision and Pattern Recognition, 2016.
[11]	Deng J, Dong W, Socher R, et al. ImageNet: A large-scale hierarchical image database[C]. Miami:Computer Vision and Pattern Recognition, 2009.
[12]	ZhouBolei, Lapedriza A, Khosla A, et al. Places:a 10 million image database for scene recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017,40(6):1452-1464. pmid: 28692961
[13]	Arandjelovic R, Zisserman A. All about VLAD[C]. Portland:Computer Vision and Pattern Recognition, 2013.
[14]	Jégou H, Douze M, Schmid C. On the burstiness of visual elements[C]. Miami:Computer Vision and Pattern Recognition, 2009.
[15]	Jégou H, Perronnin F, Douze M, et al. Aggregating local image descriptors into compact codes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012,34(9):1704-1716. pmid: 22156101
[16]	Xie L, Li F F, Liu L, et al. Improved spatial pyramid matching for scene recognition[J]. Pattern Recognition, 2018,82(3):118-129.
[17]	Li L J, Su H, Xing E P, et al. Object bank: A high-level image representation for scene classification & semantic feature sparsification[C]. Vancouver:Neural Informatino Processing Systems, 2010.
[18]	谢林, 李菲菲, 陈虬. 基于稀疏自动编码机的场景识别算法[J]. 电子科技, 2019,32(1):38-41.
	Xie Lin, Li Feifei, Chen Qiu. Scene recognition algorithm based on sparse autoencoder[J]. Electronic Science and Technology, 2019,32(1):38-41.
[19]	Wu J X, Rehg J M. CENTRIST:a visual descriptor for scene categorization[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2010,33(8):1489-501.
[20]	Lin D, Lu C W, Liao R J, et al. Learning important spatial pooling regions for scene classification[C]. Columbus:Conference on Computer Vision and Pattern Recognition,IEEE Computer Society,2014.[21] Xie Lin,Lee Feifei,Yan Yan,et al.Sparse decomposition of convolutional features for scene recognition[C].Beijing: International Conference on Computational Intelligence and Applications, 2017.
[22]	Weng Chaoqun, Wang Hongxing, Yuan Jundong, et al. Discovering class-specific spatial layouts for scene recognition[J]. IEEE Signal Processing Letters, 2017,24(8):1143-1147.
[23]	KhanS, Hayat M, Bennamoun M, et al. A discriminative representation of convolutional features for indoor scene recognition[J]. IEEE Transactions on Image Processing, 2016,25(7):3372-3383. pmid: 28113718

识别算法	识别精度/%
Object Bank^[17]	80.90
SPM^[7]	81.40
MSAE^[18]	81.97
CENTRIST^[19]	83.90
ISPR^[20]	85.10
SDCF^[21]	93.60
ResNet + Weighted Layout^[22]	94.47
DUCA^[23]	94.50
ResNet + Multi-Scale Space VLAD	94.67

基于卷积神经网络与多尺度空间编码的场景识别方法

Scene Recognition Algorithm Based on Convolutional Neural Networks and Multi-Scale space Encoding

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 6

参考文献 23

相关文章 15

Metrics

本文评价

推荐阅读 10

[1]	伍伟嘉,杨俭,袁天辰,邵志慧. 基于支持向量机的轨道结构病害识别研究[J]. 电子科技, 2022, 35(2): 27-33.
[2]	李辉,王一丞. 基于CNNCIFG-Attention模型的文本情感分类[J]. 电子科技, 2022, 35(2): 46-51.
[3]	邵志慧,杨俭,袁天辰,伍伟嘉. 基于排列熵和支持向量机的轨枕病害诊断[J]. 电子科技, 2022, 35(2): 52-58.
[4]	冯先丁,魏镜弢,吴张永,钱杰,浦友尚. 基于PCA-PSO-SVM的球磨机负荷预测研究[J]. 电子科技, 2022, 35(1): 29-34.
[5]	程晓雅,张雷. 基于深度学习的有遮挡人脸识别方法研究[J]. 电子科技, 2022, 35(1): 35-39.
[6]	战荫伟,朱百万,杨卓. 一种车脸识别算法的研究与应用[J]. 电子科技, 2021, 34(8): 1-7.
[7]	廖璐明,张伟. 基于改进VGG16网络的混合批量训练交通标志识别[J]. 电子科技, 2021, 34(8): 8-13.
[8]	马立新,豆晨飞,宋晨灿,杨天笑. 基于特征融合卷积神经网络的绝缘子无损检测[J]. 电子科技, 2021, 34(7): 26-30.
[9]	张朋飞,魏存伟,刘先康,刘安然,杨欧,林永霖. 窄带舰船目标识别方法研究[J]. 电子科技, 2021, 34(6): 50-55.
[10]	刘姝,邵杰,张颐婷,张善章. 基于双谱和谱特征的心电信号分类方法[J]. 电子科技, 2021, 34(5): 42-46.
[11]	顾伟,李菲菲,陈虬. 基于多特征融合的行人检测方法[J]. 电子科技, 2021, 34(5): 29-34.
[12]	薛永杰,巨志勇. 基于改进AlexNet的鱼类识别算法[J]. 电子科技, 2021, 34(4): 12-17.
[13]	闫书豪,乔美英. 基于一维WConv-BiLSTM的轴承故障诊断算法[J]. 电子科技, 2021, 34(4): 75-82.
[14]	葛靖,刘子龙. 基于CNN和LSTM的睡眠呼吸暂停检测算法[J]. 电子科技, 2021, 34(2): 21-26.
[15]	朱斌,刘子龙. 基于新型初始模块的卷积神经网络图像分类方法[J]. 电子科技, 2021, 34(2): 52-56.