Klasifikasi Jenis Mangga Berdasarkan Tekstur Tulang Daun Menggunakan Metode Learning Vector Quantization (LVQ)

Abstract

Mango is a fruit plants that has the potential to be developed because it has a high level of genetic diversity. Mangoes vary in shape, size and color of the fruit, indicating a fairly wide genetic diversity. Of the many genetic diversity and types of society, there are still many who cannot distinguish them. This study builds an application to distinguish mango species based on the leaf bone structure where the feature extraction process uses the Learning Vector Quantization (LVQ) method. Then do the merging to produce a specific feature vector, then the classification calculation is carried out using the Euclidean Distance method to identify the type of mango fruit. The results of the study with the amount of training data as many as 6 images and testing data as much as 15 obtained system accuracy, with the calculation results, the remaining clusters are cluster 3 with a centroid value of R = 151.67 G = 145 and B = 153.33. From the test results with 2 scenarios, the mango golek type has low values, namely 50 and 60 because the mango golek type has less brightness than the other 2 types of mango.