Laser & Optoelectronics Progress, Volume. 56, Issue 11, 111505(2019)
Camera Calibration Based on Deep Neural Network in Complex Environments
Figures & Tables(11)
Fig. 3. Loss function convergence curves under different parameter update algorithms
Table 1. Calibration errors of different methods under different shooting angles
View tableTable 1. Calibration errors of different methods under different shooting angles
Angle /(°) Error of DNN method /mm Error of SNN method /mm Error of Zhang's method / mm Avg Max Std Avg Max Std Avg Max Std 0 0.0896 0.4109 0.0553 1.8910 8.4989 1.3840 0.3887 2.2614 0.2714 15 0.0948 0.5774 0.0657 3.0027 9.1213 1.6706 0.4768 1.7119 0.3339 30 0.1271 0.7146 0.0895 3.3793 11.8804 2.1458 0.3722 3.0325 0.2733 45 0.1492 0.7831 0.0930 3.9889 15.5627 2.6222 0.2586 1.7789 0.2139 60 0.1897 0.8929 0.1003 4.6515 17.3086 2.8226 0.5350 3.2883 0.4284
Table 2. Calibration errors of different methods with highly distorted lensmm
View tableTable 2. Calibration errors of different methods with highly distorted lensmm
Error DNN method SNN method Zhang's method Avg 0.1471 7.8896 2.8714 Max 0.8990 34.2669 7.1937 Std 0.1124 4.8628 1.7055
Table 3. Calibration errors of DNN with different structuresmm
View tableTable 3. Calibration errors of DNN with different structuresmm
Error With homogeneous term Without homogeneous term Avg 0.0948 0.1325 Max 0.5774 0.8654 Std 0.0657 0.0992
Tools
Get Citation
Copy Citation Text
Peng Xiang, Bin Zhou, Yangkun Zhu, Wenkai He, Xiaogeng Yue, Yibei Tao. Camera Calibration Based on Deep Neural Network in Complex Environments[J]. Laser & Optoelectronics Progress, 2019, 56(11): 111505
Paper Information
Category: Machine Vision
Received: Dec. 20, 2018
Accepted: Jan. 7, 2019
Published Online: Jun. 13, 2019
The Author Email: Bin Zhou (seuxp@foxmail.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20