Comparison of You Only Look Once (YOLO) Algorithm Version 5 and Version 8 as Object Detection in Hilal Detection
YOLO, a deep learning-based object detection system, is suggested as a new approach to assist hilal observation to help religious authorities overcome obstacles in the rukhyat method and increase the precision of observations.
Abstract
The determination of the beginning of Ramadan is an important event for Muslims in Indonesia to start fasting. This date is determined by two stages, namely the hisab method based on the date and the rukyhat method based on hilal observation. However, the rukyhat method often faces difficulties due to many challenges such as human visual limitations. This research suggests the use of You Only Look Once (YOLO), a deep learning-based object detection system, as a new approach to assist hilal observation. The system is designed to detect the hilal accurately, even under difficult observation conditions. The application of YOLO is expected to help religious authorities overcome obstacles in the rukhyat method and increase the precision of observations. Preliminary results show that the system is able to detect the crescent moon with a high success rate, showing good potential for integrating technology into the religious domain.