Klasifikasi Jenis Mangga Berdasarkan Fitur Bentuk Dan Warna Dengan Menggunakan Metode K-Nearest Neighbor
| Main Authors: | Irnanda, Akbar Rizki, Candra, Fery |
|---|---|
| Format: | Article info application/pdf eJournal |
| Bahasa: | eng |
| Terbitan: |
Jurnal Online Mahasiswa (JOM) Bidang Teknik dan Sains
, 2018
|
| Online Access: |
https://jom.unri.ac.id/index.php/JOMFTEKNIK/article/view/22430 https://jom.unri.ac.id/index.php/JOMFTEKNIK/article/view/22430/21706 |
Daftar Isi:
- Mango is one of the fruits that popular in Indonesia. Mango has many types so that sometime people can make mistake in predicting the mango type due to similar shape and color, features among types of mango, especially Mango Arum Manis with Mango Indramayu. With an advanced technology such as MangoClassifier, predicting the type of mango can be distinguished automatically by using existing applications and K-Nearest Neighbor Method. This research aims to implement the K-Nearest Neighbor Method to detect mango types then evaluate how much the accuracy of the MangoClassifier Application is, so that it can be compared with the accuracy of other methods. This application is supported by camera equipment, matlab R2013a software, acquisition booth and K-Nearest Neighbor Method, calculates the distance between data with Euclidean distance. The acquisition booth is used to take a picture of mango object so the program will be easily to distinguish between background and object the picture. Data processing is begun with mango picture retrieval and then pre-processing the picture to get features. The used features are four, length and width from cropped object in the picture. Besides it, color and roundness of object are other features that also must be considered. This application detects four types of mango including Mango Arum Manis, Mango Indramayu, Mango Lokmay and Mango Gedong using 200 training data and 120 testing data. Accuracy obtained by this application is 92.5% so that this application is categorized as suitable to be used.Keywords : K-Nearest Neighbor, Jarak Euclidean, Matlab R2013a.