Acta Optica Sinica, Volume. 42, Issue 1, 0112005(2022)
Three-Dimensional Face Modeling Based on Multi-Scale Attention Phase Unwrapping
Figures & Tables(17)
Fig. 5. Schematic diagram of dataset construction. (a) FACE dataset; (b) MASK dataset
Fig. 6. Phase map, point cloud model and error of point cloud model obtained by different methods in undersampling experiment. (a)--(e) Phase map; (f)--(j) point cloud model; (k)--(n) error of point cloud model
Fig. 7. Error maps obtained by different methods in undersampling experiment. (a) MSAPUNet; (b) U-Net; (c) QG algorithm; (d) BC algorithm
Fig. 8. Phase maps, point cloud models and errors of point cloud model obtained by different methods in phase discontinuity experiment. (a)--(e) Phase map; (f)--(j) point cloud model; (k)--(n) error of point cloud model
Fig. 9. Error maps obtained by different methods in phase discontinuity experiment. (a) MSAPUNet; (b) U-Net; (c) QG algorithm; (d) BC algorithm
Fig. 10. Experimental results of dynamic target. (a) Texture map; (b) wrapped phase; (c) phase generated by TPU algorithm; (d) phase generated by U-Net; (e) phase generated by MSAPUNet
Table 1. RMSE and SSIM of different algorithms in FACE dataset and MASK dataset
View tableTable 1. RMSE and SSIM of different algorithms in FACE dataset and MASK dataset
Dataset QG BC U-Net MSAPUNet MRMSE /rad MSSIM MRMSE /rad MSSIM MRMSE /rad MSSIM MRMSE /rad MSSIM FACE 0.2048 0.7513 0.1888 0.7353 0.0504 0.9795 0.0387 0.9850 MASK 0.1112 0.7792 0.1031 0.8010 0.0551 0.9687 0.0273 0.9793
Table 2. Efficiency of different methods
View tableTable 2. Efficiency of different methods
Method QG BC U-Net MSAPUNet Time >10 s >2 s 30 ms 40 ms
Table 3. RMSE and SSIM of different methods in undersampling experiment
View tableTable 3. RMSE and SSIM of different methods in undersampling experiment
Index QG algorithm BC algorithm U-Net MSAPUNet MRMSE /rad 0.3392 0.2688 0.0163 0.0135 MSSIM 0.6900 0.7276 0.9952 0.9976
Table 4. Errors of point cloud model obtained by different methods in undersampling experimentunit: mm
View tableTable 4. Errors of point cloud model obtained by different methods in undersampling experimentunit: mm
Index QG algorithm BC algorithm U-Net MSAPUNet Average distance 0.7816 0.7097 0.1956 0.0006 Standard deviation 0.5569 0.5977 0.6030 0.0174
Table 5. RMSE and SSIM of different methods in phase discontinuity experiment
View tableTable 5. RMSE and SSIM of different methods in phase discontinuity experiment
Index QG algorithm BC algorithm U-Net MSAPUNet MRMSE /rad 0.1364 0.1180 0.0543 0.0291 MSSIM 0.7192 0.7888 0.9650 0.9754
Table 6. Errors of point cloud model obtained by different methods in undersampling experimentunit: mm
View tableTable 6. Errors of point cloud model obtained by different methods in undersampling experimentunit: mm
Index QG algorithm BC algorithm U-Net MSAPUNet Average distance 1.7725 2.0420 0.1083 0.0065 Standard deviation 1.1066 1.4077 0.6376 0.0915
Tools
Get Citation
Copy Citation Text
Jiangping Zhu, Ruike Wang, Zhijuan Duan, Yijie Huang, Guohuan He, Pei Zhou. Three-Dimensional Face Modeling Based on Multi-Scale Attention Phase Unwrapping[J]. Acta Optica Sinica, 2022, 42(1): 0112005
Paper Information
Category: Instrumentation, Measurement and Metrology
Received: May. 31, 2021
Accepted: Jul. 15, 2021
Published Online: Dec. 22, 2021
The Author Email: Zhou Pei (zhoupei@scu.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20