Development of automatic grading machine prototype for citrus using image processing

Citrus is one of fruits with high economic value in Indonesia. The production increases from year to year due to national fruits development program. However, postharvest equipment for citrus is very limited, leading to low technology utilization in the citrus postharvest handling. The objective of this research was to develop a real-time grading machine for citrus using an image sensor and processing unit for the quality evaluation of the citrus. The grading machine consist of a rotating fruit feeder, a belt conveyor, a color CCD camera placed in an image acquisition chamber, four openings each for a different grade of citrus, four collecting boxes to accommodate the graded citrus, a logic control panel, a computer with an image frame grabber, and a developed software to run all the installed hardware. The performance test for the grading machine then was conducted, and the result of citrus classification was observed visually to determine the performance of the machine. The testing results showed that the machine prototype could work properly, and the classification of citrus could be done based on the fruit size as conventionally done, added by the skin color of citrus as an improvement to the manual sortation.