Liu Zhen, Li Xiaojing, Li Fengjiao, Zhang Guangjun
Opt Express. 2015 Jan 12;23(1):384-400. doi: 10.1364/OE.23.000384.
Single vision sensor cannot measure an entire object because of their limited field of view. Meanwhile, multiple rigidly-fixed vision sensors for the dynamic vision measurement of three-dimensional (3D) surface profilometry are complex and sensitive to strong environmental vibrations. To overcome these problems, a novel flexible dynamic measurement method for 3D surface profilometry based on multiple vision sensors is presented in this paper. A raster binocular stereo vision sensor is combined with a wide-field camera to produce a 3D optical probe. Multiple 3D optical probes are arranged around the object being measured, then many planar targets are set up. These planar targets function as the mediator to integrate the local 3D data measured by the raster binocular stereo vision sensors into the coordinate system. The proposed method is not sensitive to strong environmental vibrations, and the positions of these 3D optical probes need not be rigidly-fixed during the measurement. The validity of the proposed method is verified in a physical experiment with two 3D optical probes. When the measuring range of raster binocular stereo vision sensor is about 0.5 m × 0.38 m × 0.4 m and the size of the measured object is about 0.7 m, the accuracy of the proposed method could reach 0.12 mm. Meanwhile, the effectiveness of the proposed method in dynamic measurement is confirmed by measuring the rotating fan blades.
单目视觉传感器由于其有限的视野无法测量整个物体。同时，用于三维（3D）表面轮廓测量动态视觉测量的多个刚性固定视觉传感器结构复杂，且对强烈的环境振动敏感。为克服这些问题，本文提出了一种基于多视觉传感器的新型3D表面轮廓测量灵活动态测量方法。将光栅双目立体视觉传感器与广角相机相结合，制成一个3D光学探头。多个3D光学探头围绕被测物体布置，然后设置许多平面靶标。这些平面靶标起到中介作用，将光栅双目立体视觉传感器测量的局部3D数据整合到坐标系中。该方法对强烈的环境振动不敏感，并且在测量过程中这些3D光学探头的位置无需刚性固定。通过使用两个3D光学探头进行物理实验验证了该方法的有效性。当光栅双目立体视觉传感器的测量范围约为0.5 m×0.38 m×0.4 m且被测物体尺寸约为0.7 m时，该方法的精度可达0.12 mm。同时，通过测量旋转的风扇叶片证实了该方法在动态测量中的有效性。
