Wang Yuwei, Wang Yajun, Liu Lu, Chen Xiangcheng
Opt Lett. 2019 Jul 1;44(13):3254-3257. doi: 10.1364/OL.44.003254.
State-of-art camera calibration methods always require sharp target images for accurate feature detection. This Letter presents a novel calibration method for a defocused camera with a conventional periodic target. Two crossed phase maps are recovered from the periodic target image by Fourier transform. Instead of intensity, feature points are extracted from the phase domain, which is robust against camera defocusing. Simulations and experiments confirm that the proposed method can realize accurate camera calibration using both focused and defocused images.
先进的相机校准方法总是需要清晰的目标图像来进行精确的特征检测。本文提出了一种针对带有传统周期性目标的散焦相机的新型校准方法。通过傅里叶变换从周期性目标图像中恢复出两个交叉相位图。从相位域中提取特征点而非强度，这对相机散焦具有鲁棒性。仿真和实验证实，所提出的方法可以使用聚焦和散焦图像实现精确的相机校准。
