Center for Optics Research and Engineering, Shandong University, Qingdao 266237, China.
School of Information Science and Engineering, Shandong University, Qingdao 266237, China.
Sensors (Basel). 2020 Jan 17;20(2):521. doi: 10.3390/s20020521.
In this paper, a real-time, dynamic three-dimensional (3D) shape reconstruction scheme based on the Fourier-transform profilometry (FTP) method is achieved with a short-wave infrared (SWIR) indium gallium arsenide (InGaAs) camera for monitoring applications in low illumination environments. A SWIR 3D shape reconstruction system is built for generating and acquiring the SWIR two-dimensional (2D) fringe pattern of the target. The depth information of the target is reconstructed by employing an improved FTP method, which has the advantages of high reconstruction accuracy and speed. The maximum error in depth for static 3D shape reconstruction is 1.15 mm for a plastic model with a maximum depth of 36 mm. Meanwhile, a real-time 3D shape reconstruction with a frame rate of 25 Hz can be realized by this system, which has great application prospects in real-time dynamic 3D shape reconstruction, such as low illumination monitoring. In addition, for real-time dynamic 3D shape reconstruction, without considering the edge areas, the maximum error in depth among all frames is 1.42 mm for a hemisphere with a depth of 35 mm, and the maximum error of the average of all frames in depth is 0.52 mm.
本文提出了一种基于傅里叶变换轮廓术（FTP）的实时动态三维（3D）形状重建方案，该方案使用短波长红外（SWIR）铟镓砷（InGaAs）相机，可用于低光照环境下的监测应用。为了生成和获取目标的 SWIR 二维（2D）条纹图案，我们构建了一个 SWIR 3D 形状重建系统。通过采用改进的 FTP 方法，可以重建目标的深度信息，该方法具有高精度和快速的优点。对于最大深度为 36mm 的塑料模型，静态 3D 形状重建的最大深度误差为 1.15mm。同时，该系统可以实现 25Hz 的实时 3D 形状重建，在低光照监测等实时动态 3D 形状重建方面具有广阔的应用前景。此外，对于实时动态 3D 形状重建，不考虑边缘区域，最大深度误差在所有帧中为 1.42mm，深度为 35mm 的半球体，所有帧深度平均值的最大误差为 0.52mm。
