Deteksi Objek pada Citra Makanan Sebagai Rekomendasi Diet Menggunakan Metode Mask R-CNN

Ratri Enggar Pawening
Meliana Eka Puteri
Agmelita Dwi Jianika
Fitriyah Hidayati

Jurnal Ilmiah Teknik Informatika dan Komunikasi
Vol. 4 No. 1 (2024): Maret : Jurnal Ilmiah Teknik Informatika dan Komunikasi

Food is one of the main needs in life for survival. This is because the energy the body needs for activities and body metabolism is obtained by consuming food. Therefore, consuming food can maintain body health and the body's metabolism can work well. In this study, the aim was to detect objects in food images, namely the types of food such as fried chicken, hamburger, seblak, baso aci, and bakwan. The method used for object detection is Mask RCNN. Previously, the image will be pre-processed, namely the resizing and annotation process. The research results show that object detection in food images has an accuracy of 72%.