Open AccessPengenalan Jenis Buah pada Citra Menggunakan Pendekatan Klasifikasi Berdasarkan Fitur Warna Lab dan Tekstur Co- Occurrence
Author(s) -
Evy Kamilah Ratnasari
Publication year - 2016
Publication title -
inform : jurnal ilmiah bidang teknologi informasi dan komunikasi/jurnal inform : jurnal ilmiah bidang teknologi informasi dan komunikasi
Language(s) - English
Resource type - Journals
eISSN - 2581-0367
pISSN - 2502-3470
DOI - 10.25139/inform.v1i2.846
Subject(s) - artificial intelligence , texture (cosmology) , rgb color model , pattern recognition (psychology) , computer science , computer vision , feature (linguistics) , image (mathematics) , mathematics , philosophy , linguistics
— Fruit recognition can be automatically applied to the field of education, industry, sales, as well as science. In the vision of computer recognition of fruit relies on four basic features that describe the characteristics of the fruit, i.e., size, color, shape, and texture. The fruit recognition through the RGB image results of cameras using the features of shape and size are not reliable and effective, because in a real data image can be composed of several different sizes of fruit on each type of fruit so it can't be identified morphologically the fruit size and uniformity that can affect the results of the classification. This journal based on the feature recognition method of building colors and textures for the classification of fruit.The classification is done by K-Nearest Neighbor based on color and texture features co-occurrence. Experimental results of 1882 dataset image of fruit for 12 different classes can recognize the fruit in both color and texture features based with the highest accuracy of 92%.