摘要: |
机器人原位测量是当前航天大型构件制造重要的发展趋势,针对大型金属构件机器人原位测量面临的非朗伯表面光学测量和弱几何微纹理表面点云匹配两大技术挑战,本文提出了一种基于光度-面结构光多传感器融合的高精度测量方法。该方法将一定数量的LED光源添加到传统条纹投影测量装置中,实现对高亮反射表面的高精度法向量估计,并将表面法向量信息与条纹投影测量的点云信息进行融合,同时设计多模态几何特征描述子提高点云匹配精度,实现测量精度的整体优化。该方法用于贮箱壁板面型测量,实验结果验证了其有效性。 |
关键词: 大型构件;原位测量;光度立体视觉;面结构光;多传感器融合 |
基金项目: |
|
Photometric and Surface Structure Light Composite Sensing Based Robot in-situ Measurement Method for Large Aerospace Component |
Hong Haibo Wan Shubiao Shen Yiping Zhong Keke
|
Shanghai Spaceflight Precision Machinery Institute, Shanghai 201600
|
Abstract: |
Robot in-situ measurement is an important development trend of current aerospace large component manufacturing. Aiming at the two technical challenges of non-Lambertian surface optical measurement and weak geometry micro texture surface point cloud matching, this paper proposes a high-precision measurement method based on photometric-surface structured light multi-sensor fusion. This method adds a certain number of LED light sources to the traditional fringe projection measurement device to achieve high-precision normal vector estimation of the highlighted reflection surface. Based on this, the surface normal vector information is fused with the point cloud information of fringe projection measurement, and a multi-modal feature descriptor is designed to improve the point cloud matching accuracy and realize the overall optimization of the measurement accuracy. The method is used to measure the surface shape of rocket tank sheet, and the experiment proves the effectiveness of the method. |
Key words: large component;in-situ measurement;photometric stereo;surface structured light;multi-sensor fusion |