Online vision-based measurement of stacks of nuclear fuel pellets in a tray

Online measurement of lengths of cylindrical nuclear fuel pellets arranged in stacks in a tray is an important prerequisite for autonomous robotic handling of pellets for the purpose of forming stacks of fixed length. This paper describes a unique measurement system that is based on dual line scan camera system calibrated with customized grids. Here, cameras acquire images for rows through synchronized triggering, while tray holding the pellet rows moves on conveyor. Subsequently, an edge crawler algorithm extracts the pellet edges and measures distance between edges after applying geometric calibration. In this paper, different sources of uncertainties along with correction methodology are described. Experiments have been conducted with the proposed system and uncertainty measurement is reported according to Guide to the Expression of Uncertainty in Measurement (GUM).