Radial Basis Function Neural Network in Identifying The Types of Mangoes

• Published in: 2020 8th International Conference on Information and Communication Technology (ICoICT) pp. 1 - 6
• Main Authors: Thaib, Faisal, Tomasila, Golda, Nivaan, Goldy Valendria, Santoso, Albertus Joko
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-06-2020
• Subjects: ANN; Artificial neural networks; classifying; Feature extraction; Gabor filters; Image color analysis; Image recognition; Mangoes; RBFNN; Shape; Training
• DOI: 10.1109/ICoICT49345.2020.9166271

Mango (Mangifera Indica L) is part of a fruit plant species that have different color and texture characteristics to indicate its type. The identification of the types of mangoes uses the manual method through direct visual observation of mangoes to be classified. At the same time, the more subjective way humans work causes differences in their determination. Therefore in the use of information technology, it is possible to classify mangoes based on their texture using a computerized system. In its completion, the acquisition process is using the camera as an image processing instrument of the recorded images. To determine the pattern of mango data taken from several samples of texture features using Gabor filters from various types of mangoes and the value of the feature extraction results through artificial neural networks (ANN). Using the Radial Base Function method, which produces weight values, is then used as a process for classifying types of mangoes. The accuracy of the test results obtained from the use of extraction methods and existing learning methods is 100%.