Yang Yuchen, Zhang Song
Appl Opt. 2024 Apr 1;63(10):2562-2569. doi: 10.1364/AO.509535.
This paper introduces a pixelwise calibration method designed for a structured light system utilizing a camera attached with a telecentric lens. In the calibration process, a white flat surface and a flat surface with circle dots serve as the calibration targets. After deriving the properties of the pinhole projector through a conventional camera calibration method using circle dots and determining the camera's attributes via 3D feature points estimation through iterative optimizations, the white surface calibration target was positioned at various poses and reconstructed with initial camera and projector calibration data. Each 3D reconstruction was fitted with a virtual ideal plane that was further used to create the pixelwise phase-to-coordinate mapping. To optimize the calibration accuracy, various angled poses of the calibration target are employed to refine the initial results. Experimental findings show that the proposed approach offers high calibration accuracy for a structured light system using a telecentric lens.
本文介绍了一种为使用配备远心镜头的相机的结构光系统设计的逐像素校准方法。在校准过程中,白色平面和带有圆点的平面用作校准目标。通过使用圆点的传统相机校准方法得出针孔投影仪的特性,并通过迭代优化通过3D特征点估计确定相机的属性后,将白色表面校准目标置于各种姿态,并使用初始相机和投影仪校准数据进行重建。每次3D重建都拟合一个虚拟理想平面,该平面进一步用于创建逐像素的相位到坐标映射。为了优化校准精度,采用校准目标的各种角度姿态来细化初始结果。实验结果表明,所提出的方法为使用远心镜头的结构光系统提供了高校准精度。