A Morphology and Coordinate Fusion-Based Positioning Method for Solar Cell Classification

The promotion of low carbonation techniques stimulates the booming demand on solar cells. One of the cutting-edge techniques is to identify online the electronic characteristics of each solar cell through the adjustment of probe positions in connection with thin wires of in-process cells for the clustering applications with different qualities. However, the fast and precise coordinate measurement of an in-process solar cell becomes a bottleneck.To this end, amorphology and coordinate fusion-based positioning method is proposed to discern accurately and efficiently the position offsets and the angle deflection of a cell. First, several image processing tools are used that reduce a cell image to an image of featured lines in a fast speed. Then, the phase of coordinate fusion takes several steps to determine the cell coordinates precisely. Experiments are conducted to evaluate the proposedmethod, where different images of in-process solar cells under both normal lights and contaminated lights are acquired by a real-world machine vision system for the estimation of the actual coordinates. The results show that the cell coordinates have been measured precisely with the maximal position error by58.89 mu m under normal lights or 90.16 mu m under contaminated lights and the maximal angle error by 0.01 degrees under normal lights or 0.25 degrees under contaminated lights. The technique is a step forward in the process of clustering solar cells associated with the cutting-edge automatic production.