• Laser & Optoelectronics Progress
  • Vol. 60, Issue 8, 0811027 (2023)
Yuan Xu, Chunyi Chen*, Xiaojuan Hu, Haiyang Yu, and Ye Tian
Author Affiliations
  • School of Computer Science and Technology, Changchun University of Science and Technology, Changchun 130022, Jilin, China
  • show less
    DOI: 10.3788/LOP230893 Cite this Article Set citation alerts
    Yuan Xu, Chunyi Chen, Xiaojuan Hu, Haiyang Yu, Ye Tian. Visual Perception Evaluation Method of Stereo Images Based on CNN-SVR[J]. Laser & Optoelectronics Progress, 2023, 60(8): 0811027 Copy Citation Text show less
    Overall framework of the objective evaluation model
    Fig. 1. Overall framework of the objective evaluation model
    Acquisition process of potential significant discomfort area. (a) Left view; (b) right view; (c) disparity map; (d) planar saliency map; (e) stereoscopic saliency map; (f) salient discomfort region
    Fig. 2. Acquisition process of potential significant discomfort area. (a) Left view; (b) right view; (c) disparity map; (d) planar saliency map; (e) stereoscopic saliency map; (f) salient discomfort region
    Part of the reconstructed images
    Fig. 3. Part of the reconstructed images
    Disparity map of potential significant discomfort area
    Fig. 4. Disparity map of potential significant discomfort area
    Regression prediction process of CNN-SVR
    Fig. 5. Regression prediction process of CNN-SVR
    Part images in the public stereo image databases
    Fig. 6. Part images in the public stereo image databases
    Scatter plot of the relationship between model-predicted value and subjective evaluation value
    Fig. 7. Scatter plot of the relationship between model-predicted value and subjective evaluation value
    ModelPLCCSROCCRMSE
    Model in Ref.[80.81280.76840.4015
    Model in Ref.[280.82400.78700.5580
    Model in Ref.[20.83100.79800.2780
    Model in Ref.[290.83580.81310.4973
    Model in Ref.[30.86680.80340.4732
    Proposed model0.87590.83350.4010
    Table 1. Performance comparison of different models on the NBU S3D-VCA dataset
    ModelPLCCSROCCRMSE
    Model in Ref.[300.73610.71900.4704
    Model in Ref.[20.82900.83000.2350
    Model in Ref.[310.83580.81310.4973
    Model in Ref.[30.93150.89760.2641
    Proposed model0.94690.90350.2547
    Table 2. Performance comparison of different models on the IVY LAB dataset
    Regression methodNBU S3D-VCAIVY LAB
    PLCCSROCCRMSEPLCCSROCCRMSE
    CNN0.78860.72840.59730.83360.80370.4375
    SVR0.86170.81620.54280.89450.86720.3560
    CNN-SVR0.87590.83350.40100.94690.90350.2547
    Table 3. Performance comparison of different regression methods
    MethodPLCCSROCCRMSE
    Global feature0.72710.63070.5994
    Local feature0.81730.76140.5042
    Proposed method0.87590.83350.4010
    Table 4. Comparison of different feature extraction methods
    Yuan Xu, Chunyi Chen, Xiaojuan Hu, Haiyang Yu, Ye Tian. Visual Perception Evaluation Method of Stereo Images Based on CNN-SVR[J]. Laser & Optoelectronics Progress, 2023, 60(8): 0811027
    Download Citation