Application of machine vision on automatic seedling transplanting

This paper presents a machine vision system for automatic seedling transplanting. To reduce transplanting time, image which seedling plants grow in tray was acquired and processed in order to identify the cells to be transplanted. Overlapping of the border seedlings and extruding leaves from neighboring cells always leads to identification failures. So a digital image processing algorithm based on morphological watersheds was developed to segment the border of leaves. The area and the perimeter of seedlings were extracted, by which whether the cell was suitable for transplanting could be determined. In this research, using tomato seedlings as samples, good identification rate for suitable seedling was obtained (98%). The result shows that this method can be used in various growing conditions of seedling, and this system can be used for automatic seedling transplanting robot.