Paper A21

Identification of Export Quality Fresh Mangosteen Based on
Area and Convex Hull Area of Fruit Calyx by Image
Processing

Usman Ahmad*1, Rudy Tjahjohutomo2, Rahmawati Nurjanah2, and Mardison3
1Department of Mechanical and Biosystem Engineering, Bogor Agricultural University, Indonesia
2Center for Agricultural Postharvest Research and Development, Bogor, Indonesia
3Center for Agricultural Engineering Research and Development, Serpong, Indonesia
Email: uahmad2010@gmail.com*

Abstract

Mangosteen is one of important fruits grown in Indonesia, and becoming an important
agricultural commodity to export. As one of a major fruit export of Indonesia, export of
mangosteen is more than 30% of the total fruits exported to several countries. Export
volume of mangosteen from Indonesia is more than 8 thousand tons, worth more than
USD 6 million per year since 2010. Mangosteen is consumed when it is fully ripe, fresh
with green calyx. For export quality, there are some additional requirements by market
such as the calyx is normal with green color. Good shape and complete calyx with green color is not connected directly to internal quality of mangosteen, however it reflects fruit freshness and more esthetic value. The objective of this research is to develop image processing algorithm to separate export quality of fresh mangosteen based on their physical appearance. After capturing image of mangosteen on a moving conveyor using a camera, some image processing techniques applied to the image using the developed algorithm. The techniques were thresholding, opening and closing, convex hull, and calyx area measurement. The results show that 91% of export quality samples were recognized correctly, and increased to 95% when convex hull operation applied to the binary image before area measurement.

Keywords: Area Measurement, Convex Hull, Image Processing, Mangosteen.

Full Text PDF of A21 Usman Ahmad et al