Defect Identification Method for Poplar Veneer Based on Progressive Growing Generated Adversarial Network and MASK R-CNN Model

被引:20
|
作者
Hu, Kai [1 ]
Wang, Baojin [1 ]
Shen, Yi [2 ]
Guan, Jieru [1 ]
Cai, Yi [1 ]
机构
[1] Nanjing Forestry Univ, Fac Mat Sci & Engn, Nanjing 210037, Peoples R China
[2] Zhenjiang Zhongfuma Machinery Co Ltd, Zhenjiang 212127, Jiangsu, Peoples R China
来源
BIORESOURCES | 2020年 / 15卷 / 02期
关键词
Veneer defects; PGGAN; MASK R-CNN; Identification; Transfer learning; WOOD; CLASSIFICATION; INSPECTION;
D O I
10.15376/biores.15.2.3041-3052
中图分类号
TB3 [工程材料学]; TS [轻工业、手工业、生活服务业];
学科分类号
0805 ; 080502 ; 0822 ;
摘要
As the main production unit of plywood, the surface defects of veneer seriously affect the quality and grade of plywood. Therefore, a new method for identifying wood defects based on progressive growing generative adversarial network (PGGAN) and the MASK R-CNN model is presented. Poplar veneer was mainly studied in this paper, and its dead knots, live knots, and insect holes were identified and classified. The PGGAN model was used to expand the dataset of wood defect images. A key ideal employed the transfer learning in the base of MASK R-CNN with a classifier layer. Lastly, the trained model was used to identify and classify the veneer defects compared with the back- propagation (BP) neural network, self-organizing map (SOM) neural network, and convolutional neural network (CNN). Experimental results showed that under the same conditions, the algorithm proposed in this paper based on PGGAN and MASK R-CNN and the model obtained through the transfer learning strategy accurately identified the defects of live knots, dead knots, and insect holes. The accuracy of identification was 99.05%, 97.05%, and 99.10%, respectively.
引用
收藏
页码:3041 / 3052
页数:12
相关论文
共 50 条
  • [41] Applied method for water-body segmentation based on mask R-CNN
    Yang, Fengyu
    Feng, Tao
    Xu, Ganyang
    Chen, Ying
    JOURNAL OF APPLIED REMOTE SENSING, 2020, 14 (01)
  • [42] Building Detection and Segmentation Method in Aerial Photograph Based on Mask R-CNN
    Zhang, Chenqiang
    Lin, Yifan
    Zhu, Hongwei
    2024 4TH INTERNATIONAL CONFERENCE ON COMPUTER COMMUNICATION AND ARTIFICIAL INTELLIGENCE, CCAI 2024, 2024, : 20 - 27
  • [43] Remote sensing image building detection method based on Mask R-CNN
    Qinzhe Han
    Qian Yin
    Xin Zheng
    Ziyi Chen
    Complex & Intelligent Systems, 2022, 8 : 1847 - 1855
  • [44] Remote sensing image building detection method based on Mask R-CNN
    Han, Qinzhe
    Yin, Qian
    Zheng, Xin
    Chen, Ziyi
    COMPLEX & INTELLIGENT SYSTEMS, 2022, 8 (03) : 1847 - 1855
  • [45] A Mask R-CNN Model with Improved Region Proposal Network for Medical Ultrasound Image
    Liu, Jun
    Li, PengFei
    INTELLIGENT COMPUTING THEORIES AND APPLICATION, PT II, 2018, 10955 : 26 - 33
  • [46] A Research on Landslides Automatic Extraction Model Based on the Improved Mask R-CNN
    Liu, Peng
    Wei, Yongming
    Wang, Qinjun
    Xie, Jingjing
    Chen, Yu
    Li, Zhichao
    Zhou, Hongying
    ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION, 2021, 10 (03)
  • [47] Pig Weight Estimation Method Based on a Framework Combining Mask R-CNN and Ensemble Regression Model
    Jiang, Sheng
    Zhang, Guoxu
    Shen, Zhencai
    Zhong, Ping
    Tan, Junyan
    Liu, Jianfeng
    ANIMALS, 2024, 14 (14):
  • [48] Research on Automatic Pavement Crack Recognition Based on the Mask R-CNN Model
    Wang, Pengcheng
    Wang, Chao
    Liu, Hongwu
    Liang, Ming
    Zheng, Wenhui
    Wang, Hao
    Zhu, Shichao
    Zhong, Guoqiang
    Liu, Shang
    COATINGS, 2023, 13 (02)
  • [49] Pest Identification and Counting of Yellow Plate in Field Based on Improved Mask R-CNN
    Rong, Minxi
    Wang, Zhizheng
    Ban, Bin
    Guo, Xiaoli
    DISCRETE DYNAMICS IN NATURE AND SOCIETY, 2022, 2022
  • [50] A Mask R-CNN based particle identification for quantitative shape evaluation of granular materials
    Yang, Dangfu
    Wang, Xiang
    Zhang, Haoran
    Yin, Zhen-yu
    Su, Dong
    Xu, Jun
    POWDER TECHNOLOGY, 2021, 392 : 296 - 305