Mage processing application for mangosteen grading with non destructive method

The amount of export quality mangosteen is only 30 - 40% of total production. This low percentage is caused by heterogeneity in fruit quality. There are some requirements in Indonesian Standard (SNI) and standard export of quality, which are normally assessed visually. Therefore image processing is recommended to be used in mangosteen grading. The objectives of this research are to determine maturity level and to make grade of mangosteen non-destructively using image processing. The Hardware used to process the image of mangosteen is the CCD camera and the computer which is equipped with video capture. Mangosteens were obtained from farmer fields in Wanayasa, Purwakarta Regency. West Java Province. The result showed that accuracy level of quality with image processing for the maturity level is 90% prediction. Moreover, accuracy level of calyx determination is 100%, and accuracy level of diameter from top view is 95.33% and 87.5% from side view.