Indonesia is a country famous for its spices wealth, spices have many benefits such as cooking and can also be used as medicine, but nowadays there are many Indonesian people who cannot distinguish each type of spices especially the rhizomes that will be used due to their shape quite similar, even though the selection of the right type of spices in accordance with the needs is very important because the spices used for cooking or medicine have different taste and efficacy, therefore the use of computer technology needs to be used to facilitate and accelerate humans in conducting classifications, this research classifies spices based on RGB and Texture colors using K-Nearest Neighbor Algorithm and distance measurement using Euclidean Distance, from 30 times the test experiment gets the result that the level of truth with K = 1 is 76%, K = 3 is equal to 67% and K = 5 by 63%. From these results it is known that based on GE colors and computer textures can classify spices but with a fairly low accuracy so that further development is needed such as adding form features. Keywords: classification, spices, K-Nearest Neighbor.
Copyrights © 2019