• Acta Optica Sinica
  • Vol. 39, Issue 4, 0415007 (2019)
Jun Yang1、*, Shun Wang2, and Peng Zhou1
Author Affiliations
  • 1 School of Electronic and Information Engineering, Lanzhou Jiaotong University, Lanzhou, Gansu 730070, China
  • 2 School of Automation and Electrical Engineering, Lanzhou Jiaotong University, Lanzhou, Gansu 730070, China
  • show less
    DOI: 10.3788/AOS201939.0415007 Cite this Article Set citation alerts
    Jun Yang, Shun Wang, Peng Zhou. Recognition and Classification for Three-Dimensional Model Based on Deep Voxel Convolution Neural Network[J]. Acta Optica Sinica, 2019, 39(4): 0415007 Copy Citation Text show less
    Voxelization of 3D mesh models. (a) Rendered models; (b) mesh models; (c) voxelization models
    Fig. 1. Voxelization of 3D mesh models. (a) Rendered models; (b) mesh models; (c) voxelization models
    Data expansion of 3D models by rotation transformation. (a) Toilet models; (b) chair models
    Fig. 2. Data expansion of 3D models by rotation transformation. (a) Toilet models; (b) chair models
    Convolution operations. (a) 2D; (b) 3D
    Fig. 3. Convolution operations. (a) 2D; (b) 3D
    Structure of convolutional neural network
    Fig. 4. Structure of convolutional neural network
    Schematic of 3D model recognition and classification
    Fig. 5. Schematic of 3D model recognition and classification
    Recognition and classification for 3D models
    Fig. 6. Recognition and classification for 3D models
    Rotation angle /(°)Accuracy rate /%
    070.6
    12077.1
    6083.5
    4087.1
    3087.7
    Table 1. Accuracy rate for recognition and classification of 3D models in expanded dataset with different rotation angles
    Kernel sizeAccuracy rate /%
    5×5×584.3
    3×3×387.7
    Table 2. Accuracy rate for recognition and classification of 3D models obtained at different sizes of convolution kernel
    ResolutionRecognition accuracy rate /%
    24×24×2481.1
    32×32×3287.7
    Table 3. Accuracy rate for recognition and classification of 3D models obtained at different resolutions
    AlgorithmRecognition accuracy rate /%
    SPH[15]68.2
    LFD[16]75.5
    3D ShapeNets[7]77.3
    Proposed algorithm87.7
    Table 4. Accuracy rate for recognition and classification of 3D models obtained with different algorithms
    Jun Yang, Shun Wang, Peng Zhou. Recognition and Classification for Three-Dimensional Model Based on Deep Voxel Convolution Neural Network[J]. Acta Optica Sinica, 2019, 39(4): 0415007
    Download Citation