<![CDATA[Pagar Alam is a region producing robusta coffee, a superior coffee variety in Indonesia with a strong taste and high caffeine quality. However, the selection process for robusta coffee beans in Pagar Alam is still traditional. It needs to be more consistent, impacting product quality, causing economic losses, and damaging the region's reputation as a producer of quality robusta coffee. Therefore, innovation is needed in the coffee bean selection process to improve the quality and competitiveness of robusta coffee from Pagar Alam. This study aims to build an image processing-based identification system for the quality of Pagar Alam robusta coffee beans. Identification is made by extracting visual features of coffee beans, including colour, shape, and size. Implementation of the Software Life Development Cycle (SDLC) through the stages of analysis, design, implementation, testing, and maintenance as a method of system development and identification process using a Support Vector Machine (SVM) with a kernel radial basis function (RBF) to extract visual features such as colour, shape, and size of coffee beans. In the system feasibility test, a percentage of 80% was obtained, a dataset of 170 data with a division ratio of 80:20, accuracy reached 91.17%, precision 100%, recall 91.17%, and F1-score 94.79%. These findings show great potential in improving the efficiency of coffee bean selection and the quality of Pagar Alam robusta coffee bean products by utilizing the support vector machine (SVM) algorithm.]]>
Copyrights © 2024
