Deteksi Bakteri pada Agar Plate Menggunakan Algoritma You Only Look Once Versi 7

Abstract

Bacteria are microorganisms that can spread diseases in various ways, one of which is through food. In Indonesia, foodborne disease frequently occurs in street vendors, packaged food, and catering services. Currently, the methods used for detecting and identifying bacteria are still performed manually, which is time-consuming and prone to errors. Therefore, a system is needed to assist experts in detecting and identifying bacteria more effectively and efficiently. This research utilizes You Only Look Once version 7 (YOLOv7) as an algorithm to detect bacteria on petri dish from species Escherichia coli, Pseudomonas aeruginosa and Staphylococcus aureus. Total image data used consists of 600 images, with 420 images as training data, 120 images as validation data, and 60 images as testing data. By implementing data augmentation, the variation of training data is increased, resulting in a total of 1.260 data. Based on the tests conducted in this research, accuracy of the system is 82,4% and mean Average Precision (mAP) value of 83,6% was obtained with the model using Adam optimizer over 200 epochs.