Haruno Sajati, Okto Dinaryanto, Arwin Datumaya Wahyudi Sumari, Astika Ayuningtyas, Shafira Ramadhani, Fatiha Eros Perdana
In the autopilot system, the aircraft must know the weather conditions in front of it through a camera to record it as a flight history or decide what steps to take in dealing with these weather conditions. In contrast to sunny and cloudy conditions, overcast conditions have two possibilities. The plane can continue its journey by penetrating the cloudy weather or avoiding it if the cloudy clouds contain lightning with too high a frequency. This study uses histogram analysis to separate the two weather conditions, cloudy and gloomy. It also separates gloomy conditions into overcast and hurricane conditions using the nearest neighbor method of the standard deviation of the data testing and data training. The program is also used to convert Red, Green, and Blue (RGB) images into grayscale and to determine the pixel value of each image test data, which is then displayed as a histogram graphical result. The results showed that the detection system could perform weather detection with the results of image test data, as many as 57 images, 45 data with successful status, and 12 data with failed status. The experiment failed because the number of pixel brightness values between the training and test data did not match the weather being tested. Based on this test, it was found that the proportion of system success was 78.95%. © 2024 IEEE.
Adisutjipto Insitut of Aerospace Technology, Faculty of Industrial Technology, Yogyakarta, Indonesia; Adisutjipto Institute of Aerospace Technology, Faculty of Aerospace Technology, Yogyakarta, Indonesia; State Polytechnic of Malang, Cognitive Artificial Intelligence Research Group (CAIRG), Department of Electrical Engineering, Malang, Indonesia