The determination of Hijri month beginnings is frequently disputed, particularly due to the subjective nature of the ru’yah (visual sighting) method, which is affected by observer bias and atmospheric variability. This study addresses a contested hilāl sighting in Aceh Jaya for Zulhijjah 1, 1446, where a photograph led to uncertainty in official determination. To mitigate misinterpretation, an artificial intelligence (AI)-based public website utilizing Google’s Teachable Machine was developed and tested. Descriptive analysis was conducted on two disputed photographic samples from the event. The AI model, trained for crescent (hilāl) image classification, consistently identified both samples as non-hilāl with a confidence exceeding 97%. The findings indicate the images were likely artifacts or noise rather than genuine crescent sightings. The research demonstrates that AI platforms provide reliable, objective support in hilāl verification, enhancing the standardization and credibility of Hijri calendar determinations for religious and scientific authorities.
Copyrights © 2025