Machine Vision and Mechanism Combination Techniques for Seedlings Quality Evaluation Based on Leaf Area

Abstract

<abstract> <bold>Abstract.</bold> Automated transplanters did seedling tray transplanting task according to seedlings quality information which was evaluated by machine vision system. Leaf area which was an important indicator of seedlings quality could obtain by processing top-view seedling images using machine vision technology. The phenomenon of leaves across cell’s rectangle and overlapping was an important factor which affected the image processing and area evaluation accuracy. In this paper, a method that combined image-processing procedure with mechanical separation was developed for the non-destructive measurement of the leaf area of seedlings in a plug tray as well as to determine seedling quality for automated transplanting. A four-step image pre-processing procedure could remove the blue separators in original RGB image and extract seedlings leaves from the background. Cucumber seedlings in booming phase was used to tested the efficiency of area calculation and quality evaluation. Compared with algorithm segmenting overlapping leaves, area calculation based on mechanical separation would be more reasonable and precise. The quality identification accuracy of methods base on mechanical separator and algorithm segmentation was 100% and 93.4%, respectively. The results showed that mechanical separator image processing procedure would be more suitable for application in an automated transplanter to distinguish the “bad” from the “good” plugs while seedlings in booming phase.