MAULANA, MUHAMAD IQBAL (2025) IMPLEMENTASI TRANSFER LEARNING MODEL ALGORITMA RESNET50V2 DALAM MENGKLASIFIKASI TINGKAT ROASTING BIJI KOPI ARABIKA. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
01 COVER.pdf Download (469kB) | Preview |
|
![]() |
Text (BAB I)
02 BAB 1.pdf Restricted to Registered users only Download (37kB) |
|
![]() |
Text (BAB II)
03 BAB 2.pdf Restricted to Registered users only Download (310kB) |
|
![]() |
Text (BAB III)
04 BAB 3.pdf Restricted to Registered users only Download (125kB) |
|
![]() |
Text (BAB IV)
05 BAB 4.pdf Restricted to Registered users only Download (405kB) |
|
![]() |
Text (BAB V)
06 BAB 5.pdf Restricted to Registered users only Download (76kB) |
|
![]() |
Text (DAFTAR PUSTAKA)
07 DAFTAR PUSTAKA.pdf Restricted to Registered users only Download (100kB) |
|
![]() |
Text (LAMPIRAN)
08 LAMPIRAN.pdf Restricted to Registered users only Download (1MB) |
Abstract
This study explores the use of transfer learning with the ResNet50V2 model to automatically classify the roasting levels of Arabica coffee beans. Roasting plays a major role in shaping the flavor and aroma of coffee, but its classification is still mostly done manually, which can be inconsistent. Using a dataset of 1,600 images categorized into four classes—green, light, medium, and dark—the model was trained in two phases: initial training and fine-tuning. Evaluation results showed excellent performance, with 99% accuracy and strong precision and recall scores. This method can assist coffee industry players, especially roasters, in achieving more consistent and efficient roasting decisions. The system also has potential for further development into mobile applications or integration with automated roasting machines, supporting the modernization of coffee production. This application of deep learning demonstrates that technology can be an accurate and efficient tool in food-related industries, particularly coffee. Keywords: transfer learning, ResNet50V2, image classification, coffee roasting, Arabica, deep learning Penelitian ini membahas penerapan metode transfer learning dengan model ResNet50V2 untuk mengklasifikasikan tingkat roasting biji kopi Arabika secara otomatis. Proses roasting punya pengaruh besar terhadap rasa dan aroma kopi, namun hingga kini sebagian besar penentuannya masih dilakukan secara manual, yang cenderung subjektif. Dengan memanfaatkan dataset berisi 1.600 gambar biji kopi pada empat kategori—green, light, medium, dan dark—model dilatih dalam dua tahap: pelatihan awal dan fine-tuning. Hasil evaluasi menunjukkan performa tinggi dengan akurasi mencapai 99%, serta nilai precision dan recall yang sangat baik. Pendekatan ini dapat membantu pelaku industri kopi, khususnya roaster, untuk meningkatkan konsistensi dan efisiensi dalam menentukan tingkat kematangan kopi. Sistem ini juga berpotensi dikembangkan lebih lanjut untuk diterapkan di perangkat mobile atau terintegrasi dengan mesin roasting otomatis, sehingga mendukung modernisasi proses produksi kopi. Penerapan deep learning seperti ini menunjukkan bahwa teknologi dapat menjadi alat bantu yang akurat dan efisien dalam industri pangan, khususnya kopi. Kata kunci: transfer learning, ResNet50V2, klasifikasi citra, roasting kopi, Arabika, deep learning
Actions (login required)
![]() |
View Item |