Analisa Perbandingan Metode Klasifikasi Euclidean Distance Dengan Fuzzy Logic Mamdani Untuk Mengidentifikasi Kematangan Buah Mangga Berdasarkan Metode Ekstraksi Fitur Ciri Statistik Orde Dua

Abstrak

Mango fruit is a fruit that has a high value in Indonesia and has a wide market share ranging from traditional markets to modern markets. At harvest time, the mango fruit is still selected manually based on visual observations from an expert on fruits. The lack of public knowledge about the maturity of this fruit makes the community sometimes difficult to choose a fruit that is ripe. So doing the research to detect mango Gedong fruit by looking at its skin texture and color. Features of Gedong will be extracted using second order feature extraction method that is energy, contrast, correlation, inverse different moment, and entropy. This application is built by comparing the two methods of classification namely Euclidean Distance, and Mamdani. By comparing the results of the two methods of classification, the user can know the maturity classification of mango Gedong more accurate. So, the result of this research shown that the most accurate method in mango fruit maturity classification is Euclidean Distance method with an accuracy of 83.33% with a total of 60 tests from 120. While Mamdani method has accuracy level under Euclidean Distance that is 63,33% with total 60 tests from 120.