Coffee is one of the most popular beverages in the world and has a different taste. One of the most famous types of coffee is arabica. This coffee has many varieties depend on the area of planting. Therefore, sometimes though both arabica coffee varieties, but may have different features such as differences in color, shape, or texture. So that sometimes farmers or coffee shop owners even people can make mistakes in recognizing arabica coffee varieties. This problem will affect the price determination of coffee, because each varieties have different prices. Therefore, needed a system that is able to recognize arabica coffee varieties accurately so it can be used as a second opinion in recognizing arabica coffee varieties. One of the technique that can be used is by imaging. This research begins with the pre-processing stage that is cropping the image manually. The next stage is feature extraction using the solidity features associated with convex hull. The last stage is classification using MultiLayer Perceptron and obtained 80% accuracy, 80% sensitivity, and 80% specificity.
Copyrights © 2018