• Laser & Optoelectronics Progress
  • Vol. 58, Issue 2, 0200003 (2021)
Yaoyao Lin1, Mei Yu1、2、*, Zhouyan He1, and Gangyi Jiang1、2
Author Affiliations
  • 1Faculty of Information Science and Engineering, Ningbo University, Ningbo, Zhejiang 315211, China
  • 2State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, Jiangsu 210093, China
  • show less
    DOI: 10.3788/LOP202158.0200003 Cite this Article Set citation alerts
    Yaoyao Lin, Mei Yu, Zhouyan He, Gangyi Jiang. Objective Quality Assessment for Three-Dimensional Meshes[J]. Laser & Optoelectronics Progress, 2021, 58(2): 0200003 Copy Citation Text show less
    Original 3D meshes of LIRIS/EPFL general-purpose. (a) Armadillo;(b) Venus;(c) Dinosaur;(d) Rockerarm
    Fig. 1. Original 3D meshes of LIRIS/EPFL general-purpose. (a) Armadillo;(b) Venus;(c) Dinosaur;(d) Rockerarm
    Original 3D meshes of LIRIS_Mask. (a) Armadillo;(b) Dinosaur;(c) Lion Head;(d) Bimba
    Fig. 2. Original 3D meshes of LIRIS_Mask. (a) Armadillo;(b) Dinosaur;(c) Lion Head;(d) Bimba
    TypePrincipleMethod
    FRGeometric distanceHD[20], RMS[21]
    Geometric LaplacianGL1[22], GL2[23]
    Network structureMSDM[14]
    Multi-scale/structureMSDM2[24]
    Dihedral angleDAME[25]
    CurvatureTPDM[26], Dong[27]
    Table 1. FR 3D mesh objective quality assessment method
    TypePrincipleMethod
    RRGlobal roughness3DWPM1[28], 3DWPM2[29]
    Local roughnessFMPD[30]
    Statistical distribution/Dihedral angleKLD Gamma[31]KLD Weibull[31]
    Table 2. RR 3D mesh objective quality assessment method
    TypePrincipleMethod
    NRStatistical distribution/Dihedral angleNR-SVR[32]
    Mean curvature/Dihedral angleNR-SVR[33]
    Dihedral angleNR-SVR[11]
    Shape index/Curvedness/Vertex scatter/AreaNR-SVR[34]
    Graph spectral features/Spatial featuresBMQA-GSES[35]
    Table 3. NR 3D mesh objective quality assessment method
    TypeMethodPLCC /%SROCC /%
    FRHD[20]11.413.8
    RMS[21]28.126.8
    GL1[22]35.533.1
    GL2[23]42.439.3
    MSDM[14]75.073.9
    MSDM2[24]81.480.4
    DAME[25]75.276.6
    TPDM[26]84.180.4
    Dong[27]87.786.6
    RR3DWPM1[28]61.869.3
    3DWPM2[29]49.649.0
    FMPD[30]83.581.9
    KLD Gamma[31]74.071.6
    KLD Weibull[31]74.171.7
    NRNR-SVR[32]87.881.5
    NR-SVR[33]82.783.6
    NR-SVR[11]86.884.6
    NR-SVR[34]87.881.9
    BMQA-GSES[35]90.587.9
    Table 4. Performance comparison of different methods on LIRIS/EPFL general-purpose database
    TypeMethodPLCC /%SROCC /%
    FRHD[20]20.226.6
    RMS[21]41.248.8
    GL1[22]39.642.0
    GL2[23]38.340.1
    MSDM[14]69.265.2
    MSDM2[24]87.389.6
    DAME[25]58.668.1
    TPDM[26]88.690.0
    Dong[27]92.692.5
    RR3DWPM1[28]31.929.4
    3DWPM2[29]42.737.4
    FMPD[30]80.880.2
    KLD Gamma[31]64.074.3
    KLD Weibull[31]74.666.6
    NRNR-SVR[32]91.290.4
    NR-SVR[33]85.488.2
    NR-SVR[11]89.191.1
    NR-SVR[34]90.160.0
    Table 5. Performance comparison of different methods on LIRIS_Mask database
    Yaoyao Lin, Mei Yu, Zhouyan He, Gangyi Jiang. Objective Quality Assessment for Three-Dimensional Meshes[J]. Laser & Optoelectronics Progress, 2021, 58(2): 0200003
    Download Citation