Laser & Optoelectronics Progress, Volume. 55, Issue 2, 021003(2018)
Depth Image Acquisition Technology Based on Improved Genetic Algorithm
Figures & Tables(7)
Fig. 3. Split results. (a) Target image; (b) OTSU method; (c) GT method; (d) GA method; (e) proposed method
Fig. 4. Comparison of experimental results. (a) Left target image;(b) right target image; (c) true depth map; (d) Ref.[17] method; (e) Ref.[18] method; (f) GA-based method; (g) proposed method
Table 1. Segmentation performance comparison of different methods
View tableTable 1. Segmentation performance comparison of different methods
Image OTSU method GT method GA method Proposed method Image threshold Operation time /ms Image threshold Operation time /ms Image threshold Operation time /ms Image threshold Operation time /ms Cones 84 18.5936 90 13.4573 95 11.5392 106 9.5967 Tsukuba 96 20.3219 104 16.7945 108 13.7329 120 10.6476
Table 2. Parameters of stereoscopic image pairs
View tableTable 2. Parameters of stereoscopic image pairs
Parameter Cones Tsukuba Image size /(pixel×pixel) 450×375 384×288 Parallax range 0.59 0.15 Display grayscale 4 16
Table 3. Performance comparison of different methods
View tableTable 3. Performance comparison of different methods
Method Cones Tsukuba σ DSS σ DSS Proposed method 0.2540 1.2825×109 0.0360 2.3406×108 Ref.[17] method 0.3727 1.8818×109 0.0831 5.3952×108 Ref.[18] method 0.4012 2.0302×109 0.0968 6.2936×108 GA-based method 0.3285 1.6554×109 0.0684 4.4471×108
Tools
Get Citation
Copy Citation Text
Qi Wang, Yan Piao. Depth Image Acquisition Technology Based on Improved Genetic Algorithm[J]. Laser & Optoelectronics Progress, 2018, 55(2): 021003
Paper Information
Category: Image processing
Received: Jul. 30, 2017
Accepted: --
Published Online: Sep. 10, 2018
The Author Email: Yan Piao (piaoyan@cust.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20