留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于卷积神经网络的食品图像识别

廖恩红 李会芳 王华 庞雄文

廖恩红, 李会芳, 王华, 庞雄文. 基于卷积神经网络的食品图像识别[J]. 华南师范大学学报(自然科学版), 2019, 51(4): 113-119. doi: 10.6054/j.jscnun.2019075
引用本文: 廖恩红, 李会芳, 王华, 庞雄文. 基于卷积神经网络的食品图像识别[J]. 华南师范大学学报(自然科学版), 2019, 51(4): 113-119. doi: 10.6054/j.jscnun.2019075
LIAO Enhong, LI Huifang, WANG Hua, PANG Xiongwen. Food Image Recognition Based on Convolutional Neural Network[J]. Journal of South China normal University (Natural Science Edition), 2019, 51(4): 113-119. doi: 10.6054/j.jscnun.2019075
Citation: LIAO Enhong, LI Huifang, WANG Hua, PANG Xiongwen. Food Image Recognition Based on Convolutional Neural Network[J]. Journal of South China normal University (Natural Science Edition), 2019, 51(4): 113-119. doi: 10.6054/j.jscnun.2019075

基于卷积神经网络的食品图像识别

doi: 10.6054/j.jscnun.2019075
基金项目: 

广东省科技计划项目 2017B010126002

广东省科技计划项目 2017A010101008

广东省科技计划项目 2017A010101014

详细信息
    通讯作者:

    庞雄文, 副教授, Email:augepang@163.com

  • 中图分类号: TP183

Food Image Recognition Based on Convolutional Neural Network

  • 摘要: 针对传统食品图像识别方法提取特征能力差、准确率差、运行效率差和卷积神经网络识别相似食品图像难度大等问题, 提出了一种新的食品图像识别模型ChinaFood-CNN, 以实现对食物的精准分类; 在多分类损失函数SoftmaxWithLoss的基础上, 针对食品图像类间相似性大的问题, 提出了最大类间距损失函数(MCSWithLoss), 以增大相似类之间的距离, 实现相似类的区分; 针对随机选取样本时的训练集冗余问题, 在网络模型训练时使用负样本选择算法.实验结果表明:ChinaFood-CNN模型对食品图像的识别准确率达69.2%, 分别比AlexNet、VGG16、ResNet模型提升了17.6%、16.8%和3.6%.
  • 图  1  ChinaFood-CNN模型结构

    Figure  1.  The structure of ChinaFood-CNN model

    图  2  4种学习率衰减方法在ChinaFood482数据集上的准确率和loss值

    Figure  2.  The accuracy and loss value of four learning rate attenuation methods in ChinaFood482 Dataset

    图  3  4种模型在UEC-256和ChinaFood482食品数据集上的准确率

    Figure  3.  The accuracy of four network models in UEC-256 and ChinaFood482

    图  4  SoftmaxWithLoss分类效果

    Figure  4.  The classification effect of Softmax WithLoss

    图  5  不同参数的MCS_SoftmaxWithLoss分类效果

    Figure  5.  The classification effect of MCS_SoftmaxWithLoss with different parameters

    图  6  不同损失函数的准确率

    Figure  6.  The accuracy of different loss functions

    图  7  相似类别图片组Ⅰ的测试结果

    Figure  7.  Test results of similar category picture group Ⅰ

    图  8  相似类别图片组Ⅱ的测试结果

    Figure  8.  Test results of similar category picture group Ⅱ

    图  9  不同特征提取模块的Faster R-CNN模型的识别结果

    Figure  9.  The recognition result of different feature extraction modules of Faster R-CNN

    图  10  不同特征提取模块的Faster R-CNN模型的识别准确率

    Figure  10.  The recognition accuracy of different feature extraction modules of Faster R-CNN

  • [1] LOWE D G. Object recognition from local scale-invariant features[C]//Proceedings of the 7th IEEE International Conference on Computer Vision. Kerkyra: IEEE, 1999: 1150-1157.
    [2] DALAL N, TRIGGS B. Histograms of oriented gradients for human detection[C]//Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Diego: IEEE, 2005: 886-893.
    [3] HOASHI H, JOUTOU T, YANAI K. Image recognition of 85 food categories by feature fusion[C]//Proceedings of the 2010 IEEE International Symposium on Multimedia. Singapore: IEEE, 2010: 296-301.
    [4] JOUTOU T, YANAI K. A food recognition system with multiple kernel learning[C]//Proceedings of 16th IEEE International Conference on Image Processing. Cairo: IEEE, 2009: 285-288.
    [5] CORTES C, VAPNIK V. Support-vector networks[J]. Machine Learning, 1995, 20(3):273-297. http://d.old.wanfangdata.com.cn/Periodical/hwyhmb200803006
    [6] LIU C, CAO Y, LUO Y, et al. Deepfood: deep learning-based food image recognition for computer-aided dietary assessment[C]//Proceedings of International Conference on Smart Homes and Health Telematics. Berlin: Springer, 2016: 37-48.
    [7] MEZGEC S, SELJAK B K. NutriNet:a deep learning food and drink image recognition system for dietary assessment[J]. Nutrients, 2017, 9(7):E657/1-19. http://d.old.wanfangdata.com.cn/Periodical/dlyj200502007
    [8] CIOCCA G, NAPOLETANO P, SCHETTINI R. Food recognition:a new dataset, experiments, and results[J]. IEEE Journal of Biomedical and Health Informatics, 2016, 21(3):588-598. http://d.old.wanfangdata.com.cn/NSTLQK/NSTL_QKJJ0218788487/
    [9] AGUILAR E, REMESEIRO B, BOLAÑOS M, et al. Grab, pay, and eat:semantic food detection for smart restaurants[J]. IEEE Transactions on Multimedia, 2018, 20(12):3266-3275. doi: 10.1109/TMM.2018.2831627
    [10] CHRISTODOULIDIS S, ANTHIMOPOULOS M, MOUGIA-KAKOU S. Food recognition for dietary assessment using deep convolutional neural networks[C]//Proceedings of International Conference on Image Analysis and Proce-ssing. Berlin: Springer, 2015: 458-465.
    [11] POULADZADEH P. A cloud-assisted mobile food recognition system[D]. Ottawa: University of Ottawa, 2017.
    [12] MARTINEL N, FORESTI G L, MICHELONI C. Wide-slice residual networks for food recognition[C]//Proceedings of 2018 IEEE Winter Conference on Applications of Computer Vision. Lake Tahoe: IEEE, 2018: 567-576.
    [13] 罗建豪, 吴建鑫.基于深度卷积特征的细粒度图像分类研究综述[J].自动化学报, 2017, 43(8):1306-1318. http://d.old.wanfangdata.com.cn/Periodical/zdhxb201708002

    LUO J H, WU J X. A review of fine-grained image classification based on deep convolution[J]. Acta Automatica Sinica, 2017, 43(8):1306-1318. http://d.old.wanfangdata.com.cn/Periodical/zdhxb201708002
    [14] CHEN M, DHINGRA K, WU W, et al. PFID: Pittsburgh fast-food image dataset[C]//Proceedings of the 16th IEEE International Conference on Image Processing. Cairo: IEEE, 2009: 289-292.
    [15] BOSSARD L, GUILLAUMIN M, VAN GOOL L. Food-101-mining discriminative components with random fo-rests[C]//Proceedings of European Conference on Computer Vision. Berlin: Springer, 2014: 446-461.
    [16] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. System and method for parallelizing convolutional neural networks: US9563840[P]. 2017-02-07.
    [17] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[J/OL]. (2014-04-10)[2019-04-13]. https://arxiv.org/abs/1409.1556.
    [18] HE K, ZHANG X Y, REN S Q, et al. Deep residual lear-ning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.
    [19] TAO Y, ZHANG D P, HU K, et al. Multi-pose face recognition by dynamic loss weights[C]//Proceedings of 4th International Conference on Systems and Informatics. Hangzhou: IEEE, 2017: 453-458.
    [20] SCHROFF F, KALENICHENKO D, PHILBIN J. Facenet: a unified embedding for face recognition and clustering[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Boston: IEEE, 2015: 815-823.
    [21] JIA Y Q, SHELHAMER E, DONAHUE J, et al. Caffe: convolutional architecture for fast feature embedding[C]//Proceedings of the 22nd ACM International Conference on Multimedia. New York: ACM, 2014: 675-678.
    [22] REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN:towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2017, 39(6):1137-1149.
  • 加载中
图(10)
计量
  • 文章访问数:  2680
  • HTML全文浏览量:  1405
  • PDF下载量:  129
  • 被引次数: 0
出版历程
  • 收稿日期:  2019-05-17
  • 刊出日期:  2019-08-25

目录

    /

    返回文章
    返回