Texture and Materials Image Classification Based on Wavelet Pooling Layer in CNN

被引:6
|
作者
Manuel Fortuna-Cervantes, Juan [1 ]
Tulio Ramirez-Torres, Marco [2 ]
Mejia-Carlos, Marcela [1 ]
Salome Murguia, Jose [3 ,4 ]
Martinez-Carranza, Jose [5 ]
Soubervielle-Montalvo, Carlos [6 ]
Arturo Guerra-Garcia, Cesar [2 ]
机构
[1] Univ Autonoma San Luis Potosi, Inst Invest Comunicac Opt, Alvaro Obregon 64, San Luis Potosi 78000, San Luis Potosi, Mexico
[2] Univ Autonoma San Luis Potosi, Coordinac Acad Reg Altiplano Oeste, Carretera Salinas Santo Domingo 200 Salinas, San Luis Potosi 78600, San Luis Potosi, Mexico
[3] Univ Autonoma San Luis Potosi, Fac Ciencias, Av Chapultepec 1570, San Luis Potosi 78295, San Luis Potosi, Mexico
[4] Univ Autonoma San Luis Potosi, Fac Ciencias, Lab Nacl CI3M, Av Chapultepec 1570, San Luis Potosi 78295, San Luis Potosi, Mexico
[5] Inst Nacl Astrofis Opt & Elect INAOE, Dept Computat Sci, Puebla 72840, Mexico
[6] Fac Ingn UASLP, Ctr Invest & Estudios Posgrad, Av Dr Manuel Nava 8, San Luis Potosi 78290, San Luis Potosi, Mexico
来源
APPLIED SCIENCES-BASEL | 2022年 / 12卷 / 07期
基金
芬兰科学院;
关键词
texture and materials classification; CNNs; wavelet pooling layer; CONVOLUTIONAL NEURAL-NETWORKS;
D O I
10.3390/app12073592
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
Convolutional Neural Networks (CNNs) have recently been proposed as a solution in texture and material classification in computer vision. However, inside CNNs, the internal layers of pooling often cause a loss of information and, therefore, is detrimental to learning the architecture. Moreover, when considering images with repetitive and essential patterns, the loss of this information affects the performance of subsequent stages, such as feature extraction and analysis. In this paper, to solve this problem, we propose a classification system with a new pooling method called Discrete Wavelet Transform Pooling (DWTP). This method is based on the image decomposition into sub-bands, in which the first level sub-band is considered as its output. The objective is to obtain approximation and detail information. As a result, this information can be concatenated in different combinations. In addition, wavelet pooling uses wavelets to reduce the size of the feature map. Combining these methods provides acceptable classification performance for three databases (CIFAR-10, DTD, and FMD). We argue that this helps eliminate overfitting and that the learning graphs reflect that the datasets show learning generalization. Therefore, our results indicate that our method based on wavelet analysis is feasible for texture and material classification. Moreover, in some cases, it outperforms traditional methods.
引用
收藏
页数:26
相关论文
共 50 条
  • [1] Image texture classification based on wavelet transform and SVM
    Lan, Gao
    Song Qingguo
    Chuang, Li
    Qing, Hua
    Chuang, Yang
    [J]. PROCEEDINGS OF THE INTERNATIONAL CONFERENCE INFORMATION COMPUTING AND AUTOMATION, VOLS 1-3, 2008, : 834 - +
  • [2] Texture CNN for Histopathological Image Classification
    de Matos, Jonathan
    Britto, Alceu de S., Jr.
    de Oliveira, Luiz E. S.
    Koerich, Alessandro L.
    [J]. 2019 IEEE 32ND INTERNATIONAL SYMPOSIUM ON COMPUTER-BASED MEDICAL SYSTEMS (CBMS), 2019, : 580 - 583
  • [3] Image descreening by GA-CNN-based texture classification
    Shou, YW
    Lin, CT
    [J]. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS, 2004, 51 (11) : 2287 - 2299
  • [4] Wavelet-based texture analysis for SAR image classification
    Thitimajshima, P
    [J]. APPLICATIONS OF DIGITAL IMAGE PROCESSING XXII, 1999, 3808 : 717 - 720
  • [5] Wavelet-Attention CNN for image classification
    Zhao, Xiangyu
    Huang, Peng
    Shu, Xiangbo
    [J]. MULTIMEDIA SYSTEMS, 2022, 28 (03) : 915 - 924
  • [6] Wavelet-Attention CNN for image classification
    Xiangyu Zhao
    Peng Huang
    Xiangbo Shu
    [J]. Multimedia Systems, 2022, 28 : 915 - 924
  • [7] Texture image segmentation method based on multi-layer CNN
    Liu, GX
    Oe, S
    [J]. 12TH IEEE INTERNATIONAL CONFERENCE ON TOOLS WITH ARTIFICIAL INTELLIGENCE, PROCEEDINGS, 2000, : 147 - 150
  • [8] Wavelet based texture classification
    Sebe, N
    Lew, MS
    [J]. 15TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION, VOL 3, PROCEEDINGS: IMAGE, SPEECH AND SIGNAL PROCESSING, 2000, : 947 - 950
  • [9] ViP: Virtual Pooling for Accelerating CNN-based Image Classification and Object Detection
    Chen, Zhuo
    Zhang, Jiyuan
    Ding, Ruizhou
    Marculescu, Diana
    [J]. 2020 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV), 2020, : 1169 - 1178
  • [10] Graph wavelet transform for image texture classification
    Qiao, Yu-Long
    Zhao, Yue
    Song, Chun-Yan
    Zhang, Kai-Ge
    Xiang, Xue-Zhi
    [J]. IET IMAGE PROCESSING, 2021, 15 (10) : 2372 - 2383