Increasing Throughput with an Embedded Vision System


A large agricultural corporation was seeking a better way to inspect a tray of 600 seeds. Each seed in the tray needed to be analyzed individually and identically, but current self-developed methods were heavily reliant on human input. These methods were too slow and results varied from inspector to inspector.


Our solution was multi-faceted in nature and driven with an embedded vision system. Three individual, high resolution cameras using multiple lighting solutions image the entire tray and provide feedback almost instantly. Trays and tray wells are analyzed individually by self-adjusting inspection criteria as the inspection region scans trays under multiple cameras.