New crescent moon detection using Circular Hough Transform (CHT)

• Published in: Astronomy and computing 2025-04, Vol.51, p.100902, Article 100902
• Main Authors: Nasir, A.L.A.M., Umar, R., Yussof, W.N.J.W., Ahmad, N., Zulkeflee, A.N., Sabri, N.H., Chik, W.M.Y.W., Mahiddin, N.A., Bely, A.H., Izdihar, A.I.S.
• Format: Article
• Language: English
• Subjects: CHT; Image processing; New crescent moon; OpenCV
• ISSN: 2213-1337
• DOI: 10.1016/j.ascom.2024.100902

Over 1.5 billion Muslims worldwide rely on a calendar based on the observation of the new crescent moon, particularly during the Hijri calendar months of Ramadhan, Shawal, and Zulhijjah. Accurately detecting the new crescent moon immediately after its conjunction is essential for ensuring the precision of the calendar, but it poses a technical challenge due to the low contrast between the new crescent moon and its background. This study aimed to solve the challenge of reliably identifying the new crescent moon at this critical phase, which is vital for accurately determining important dates that signify religious events. It intended to create and verify a computer vision model that would greatly improves the precision and efficiency of identifying a new crescent moon based on images. This will assist in precisely identifying important dates relevant to Muslim religious practices. The methodology employs image preprocessing techniques such as Gaussian Blur for image smoothing and Adaptive Thresholding for contrast enhancement to optimize the visibility of the crescent moon in a single image. The Circular Hough Transform (CHT) technique was used to accurately detect the new crescent moon, while the OpenCV software package helped to implement all of these methods, thus, providing a strong foundation for precisely identifying the new crescent moon. Findings of this study, which involved observations at Kolej Ugama Sultan Zainal Abidin (KUSZA) Observatory, Teluk Kemang Observatory, and Miri Observatory, indicate that the utilization of image processing techniques results in significant improvements in both the efficiency and precision of identifying and detecting the new crescent moon. These enhancements exceed the performance of conventional digital imaging methods employed during regular observations. The model’s processing rate surpasses the usual observation efficiency and showcases exceptional accuracy in distinguishing images that feature the new crescent moon from those that do not. In conclusion, the proposed model demonstrates an effective and efficient methodology for improving the precision of lunar calendar management and the quick planning of important Islamic religious events in accordance with the lunar cycle. This technological advancement offers a reliable method for accurately seeing the crescent moon, which has important implications for enhancing the precision of lunar-based calendars and coordinating important Islamic religious occasions.