Fardana, Freda Adi (2025) Estimasi Kandungan Gizi Makanan Indonesia Berdasarkan Deteksi Objek Menggunakan YOLOv8 dan XGBoost. Undergraduate thesis, Politeknik Negeri Jember.
![[img]](https://sipora.polije.ac.id/style/images/fileicons/text.png) |
Text (Abstract)
Abstract.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (229kB) |
|
Text (Bab 1 Pendahuluan)
Bab 1 Pendahuluan.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (230kB) |
|
Text (Daftar Pustaka)
Daftar Pustaka.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (192kB) |
|
Text (Laporan Lengkap)
Skripsi - SIM.pdf - Submitted Version Restricted to Registered users only Download (4MB) | Request a copy |
Abstract
Nutrition plays an important role in growth and development, yet nutritional imbalances still frequently occur due to a lack of public knowledge. This study developed a system for food type detection and nutrition estimation using the YOLOv8 and XGBoost methods, implemented through a Telegram bot. Experimental results show that YOLOv8 achieves very high accuracy in container detection, reaching 95–97% and 91–97% in scenarios with one or two containers per image, respectively, but experiences a decrease to 27% accuracy in the threecontainer scenario, especially for oil paper containers, with an average accuracy of 88.1%. For segmentation, the model can recognize objects such as fried rice, fried chicken, fried tofu, fried tempeh, and fried egg with accuracies above 92%, but experiences decreased accuracy for objects with irregular shapes or low contrast against the background. The average segmentation accuracy is 91.22% in scenarios where objects are combined in a single container. For weight prediction, the XGBoost model achieved an MAE of 21.138 grams, an RMSE of approximately 50.10 grams, and an R² value of 0.7838, with the largest errors found in fried rice (-155.97 grams) and fried chicken (+17.31 grams) due to variations in shape and food composition.
| Item Type: | Thesis (Undergraduate) | ||||||
|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||
| Uncontrolled Keywords: | deteksi makanan, YOLOv8, segmentasi, XGBoost, estimasi gizi | ||||||
| Subjects: | 410 - Rumpun Ilmu Teknik > 450 - Teknik Elektro dan Informatika > 458 - Teknik Informatika | ||||||
| Divisions: | Jurusan Teknologi Informasi > Prodi D4 Teknik Informatika > Tugas Akhir | ||||||
| Depositing User: | Freda Adi Fardana | ||||||
| Date Deposited: | 26 Jun 2025 02:48 | ||||||
| Last Modified: | 26 Jun 2025 02:49 | ||||||
| URI: | https://sipora.polije.ac.id/id/eprint/42573 |
Actions (login required)
 |
View Item |
