EKSTRKASI FITUR BENTUK MENGGUNAKAN METODE CONVEX HULL UNTUK KLASIFIKASI JENIS PISANG MENGGUNAKAN K-NEAREST NEIGHBOR

Abstract

There are many types of bananas in Indonesia, for example Ambon bananas, Kepok bananas, Susu bananas, Mas bananas and Cavendish bananas. With the existence of many types of bananas, to determine the type of banana, it still uses judgment by the human eye based on the shape of the banana. In this study to find out the type of banana using automatic classification of banana image input. The results of the classification are expected to determine the type of banana based on tranning data. Banana image input is extracted using the Convex Hull method to obtain Solidity and Convexity values. The steps to get the classification value are carried out by inputting the banana image, converting the color to binary (black and white), extracting the Convex Hull shape feature, calculating the convexity Solidity value and the Solidity and convexity value of the banana image, the classification process is carried out using the K-Nearst Neighbor method ( K-NN). To determine the success rate of the classification results carried out the testing process. From the test results, the process of calculating the accuracy of all the data tested is carried out. The results obtained in this study with an accuracy value of 56%.