将双目传感器和线结构光有机地结合到一起, 设计了一套能够对大尺寸被测物进行无导轨三维扫描的测量系统。通过标定实现双目传感器与结构光传感器坐标系的统一。提出了一种双线投影模型, 实现了双目标记点的立体匹配和标记点传感器坐标的测量。设计了基于线面约束的结构光测量模型, 实现了物体表面光条三维信息的获取。采用空间几何相对不变性原则实现标记点传感器坐标和世界坐标的全局匹配, 进而确定在测量位置处传感器坐标系与世界坐标系的转换关系。最终将传感器实时扫描测量的物体表面的三维点云数据转换到世界坐标系下, 实现对被测物三维的扫描测量, 工作距离下扫描测量精度优于0.08 mm。

A system about 3D scanning measurement towards large objects without rails was developed in this paper, which combined binocular sensor with structured-light sensor together and achieved the unity of their coordinate systems through calibration. The double-line projection model was presented, through which the system can complete the stereo matching of the mark points in the binocular sensor and get mark points′ 3D coordinates. Based on the double-line projection model, a structured-light model named line-plane constraint was designed to get the modulated information of the light strips on the objects′ surface. In order to build the relationship between the sensor coordinate system and the global coordinate system, the property of the space geometrical invariability of the mark points was used. In this way, the entire system of the sensor can convert the 3D points cloud data measured at different locations to the global system in real time. Thus, the entire sensor system can complete the 3D scanning measurement of the object, the measurement precision was better than 0.08 mm within the working distance.