VISION INTELLIGENCE AND SOFTWARE CONTROL FOR THE AUTOMATION OF PLANT MICROPROPAGATION

This paper describes a vision-directed robot system which is designed to automate several aspects of plant micropropagation. Viable microplants, growing in a lawn of specimens under cultivation, are identified in terms of their visual characteristics by the vision sub-system. The robot transfers candidate microplants to individual containers for further growth. Non-viable specimens are left behind and later discarded. Microplant image processing and analysis are discussed together with robot control techniques. A user-friendly software operating environment, developed to manipulate the vision and robot sub-systems, is outlined. It allows users to edit, debug and execute their own simple image processing and robot control procedures and permits programme designers to modify and debug image processing and robot control programs.