A NEW LUNAR OBSERVATIONAL TECHNIQUE USING ARTIFICIAL INTELLIGENT SYSTEM

• Main Author: Muztaba, Robiatul
• Format: Dissertations
• Language: Indonesia
• Subjects: Astronomi
• Online Access: https://digilib.itb.ac.id/gdl/view/79841
• Institution: Institut Teknologi Bandung

Observation of the crescent or so-called Hilal is still a fascinating issue to continue to be developed. In Indonesia, where the majority of the population is Muslim, this is very important, especially in determining the beginning of the Qamariyah month in particular, Ramadan, Shawwal and Dzulhijjah. The heterogeneities in weather condition, climate, and observation locations cause different human perception of new moon visibility. Usually, to increase the chances of observing the Hilal we have to look for locations where west horizon sky free of obstacles, such as beaches and mountain peaks. But in fact, observing the crescent in these places is still very difficult. First, the different manner of methods and techniques adopted by users in using the telescope. Second, the problem of the contrast between the crescent and sky background. As technology advances, crescent moon observations take advantage of image processing methods. Image processing aims to enhance quality the output image to be better than the input image so that the crescent moon can be easy to analyze. However, another problem arises when the results of image processing cannot be accepted because they are not in accordance with the provisions of jurisprudence and the opinion of religious experts. This problem arises when there is still a significant time lag between the recording and image processing, causing a sense of skepticism (suspicion or distrust of the image processing results) so that it cannot be used as a basis for determining the entry of the start of a new month. Therefore, this study provides technological updates to develop a crescent moon observation system by implementing an artificial intelligence system program, namely computer vision. The system acts as a decision support system that uses computer vision to support a robotic telescope so that it can detect the crescent moon more quickly and precisely.