PENGOLAHAN CITRA UNTUK IDENTIFIKASI KEMATANGAN BUAH JERUK DENGAN MENGGUNAKAN METODE BACKPROPAGATION BERDASARKAN NILAI HSV

Abstract

Tanah Karo is one of the mountainous areas that has very cool air, so Tanah Karo is a producer of citrus fruits that produce citrus fruits of good quality. But unfortunately the citrus farmers in the area still use conventional (manual) methods or by seeing with the human eye's eyes in selecting citrus fruits of suitable maturity and service without special knowledge and only from their experience. However, human vision has the limitation that the human eye will experience fatigue. Harvesting unripe citrus fruits results in inappropriate quality of citrus fruit being marketed and if you harvest too ripe citrus fruits, it will cause the citrus fruits to rot quickly when they are distributed to agents or buyers in the market. From the input pattern / image of citrus fruits as training data and training targets, it can identify the ripeness of citrus fruits using the backpropagation method. Based on the citrus fruit image data, it can recognize the pattern of citrus fruit that has a maturity level that matches the digital image using the backpropagation method, with the accuracy of training and testing data being 100%.