Acta Optica Sinica, Volume. 38, Issue 9, 0915004(2018)
Depth Map Generation Algorithm Based on Visual Dictionary
Figures & Tables(7)
Fig. 3. Sample image Gaussian pyramid patches after data cleaning. (a) Layer1, octave1, variance is σ;(b) layer2, octave1, variance is 2σ; (c) layer1, octave2, variance is 2σ; (d) layer2, octave2, variance is 22σ;(e) layer1, octave3, variance is 4σ; (f) layer2, octave3, variance is 42σ
Fig. 5. Depth maps generated by four algorithms. (a) 2D color images; (b) ground truth depth maps; (c) Ref. [2] algorithm; (d) Ref. [14] algorithm; (e) Ref. [16] algorithm;(f) proposed algorithm
Table 1. PSNR values of depth maps generated by four algorithmsdB
View tableTable 1. PSNR values of depth maps generated by four algorithmsdB
Image Ref. [2] algorithm Ref. [14] algorithm Ref. [16] algorithm Proposed algorithm Image 1 26.05 25.05 25.45 28.28 Image 2 25.00 24.97 25.05 25.54 Image 3 29.94 29.96 31.35 31.43 Image 4 26.41 26.64 26.85 27.98
Table 2. Results comparison of depth maps generated by four algorithms
View tableTable 2. Results comparison of depth maps generated by four algorithms
Parameter Ref. [2] algorithm Ref. [14] algorithm Ref. [16] algorithm Proposed algorithm PSNR /dB 26.48 26.36 26.20 28.36 Rel Err 0.74 0.49 0.72 0.32 RMSE 12.10 12.26 12.49 9.47
Tools
Get Citation
Copy Citation Text
Jieping Liu, Huasheng Zhou, Langheng Yu, Shuhao Ding, Yaling Liang. Depth Map Generation Algorithm Based on Visual Dictionary[J]. Acta Optica Sinica, 2018, 38(9): 0915004
Paper Information
Category: Machine Vision
Received: Dec. 13, 2017
Accepted: Apr. 28, 2018
Published Online: May. 9, 2019
The Author Email:
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20