Pengembangan Metode Deteksi Tingkat Kematangan Melon Varietas Cantaloupe Menggunakan Pengolahan Citra

Abstract

Until now, determination of melon fruit harvest time which is done thrugh manual observation result in harvested fruits with low accuracy and consistency in ripeness, due to human visual limitation. Therefore, technology such as image processing is needed to replace human visual system, with ability to detect ripeness level of melon. The objectives of this research is to study and determine correlation of ripeness parameters analyzed by image processing and those measured directly. The results show that there are correlation between fruits harvested at 54 DAP (days after planting) and those harvested at 60 DAP, but not significant. The correlation of the two measurement methods is very low (r<0.5). However, using discriminant analysis, the two groups of harvested melons can be distinguished with 68.4% of successfull rate. The image parameter that can be used to distingusih the two groups of fruits is red color index of the fruit skin.