Development of an automatic grading system for green hawthorn leaf using image processing

In the course of grading green hawthorn leaves, the dependence degree of manpower and production price are all high, and there is not quantitative size standard by hand. So this study was conducted to develop and automatic grading system for green hawthorn leaves. The grading system consisted of a control computer, a colour CCD camera(WV-CP450) for image processing, an image processing board (DT3153), an A/D board (PCI-9111 series) for controlling movement of conveyor belt, and two conveyor belts for implementing operation of grading. According to area size of green hawthorn leaves with the clustering analysis theory, the size was classified into 3 groups: large size was above 120 cm², middle size was 90~120 cm², and small size below 90 cm². Grading accuracies for green hawthorn leaves were 90.2%, grading capacity of the system was 1200 pieces/hour. It can also meet the need of the industrial production well, and enhance the autoimmunization and intelligentization of impurity removing from hawthorn leaves greatly.