Task-Specific Uncertainty for Point Cloud Measurements

In this research, we have developed a data fusion technique to integrate the measurement point clouds collected from a workpiece. One dataset was obtained using a high-accuracy, low-density, and low-throughput tool, and a low-accuracy, high-density, and high-throughput instrument generated the other dataset. By fusing these two datasets we were able to achieve a high-accuracy and high-density point cloud of the workpiece. This will speed up the required time needed to measure the industrial parts at scale. To develop this model, we first generated a 1D simulation as a proof of concept for our study. Next, this model was expanded to a 2D simulation, and performed the fusion test on a workpiece surface. Then, the fusion model was applied to a cylindrical shape workpiece in 3D. Our results suggest that the fusion model that we have developed in this work can be utilized at various scales to make high-density, high-accuracy, and high-throughput measurements of industrial parts feasible.