| dc.description.abstract | Coffee has an important role in international trade. The price of coffee beans depends on their quality, which has a direct correlation with the taste of the coffee. There are provisions for determining the quality of coffee beans based on price. The quality can be detected based on the color and texture of the coffee beans. This research produces a system that can detect the quality of coffee beans by looking at color and texture. The quality of the coffee beans uses the Faster Regional Convolutional Neural Network (Faster R-CNN) method. then divided into three levels, namely quality A, quality B, and quality C. Quality A in this study is a type of coffee bean with a high selling price, quality B is a type of coffee bean with a medium selling price and quality C is a type of coffee bean with a low selling price. The data used in this research amounted to 600 data which was then divided into 420 training data, 60 validation data and 120 testing data. After testing, this research resulted in an accuracy of 91.6%. Based on the accuracy values obtained, it can be concluded that the system built using the Faster Regional Convolutional Neural Network (Faster R-CNN) method is good at detecting the quality of coffee beans. | en_US |
