A low-cost local cloud monitoring system

Abstract

In this paper, a local cloud monitoring station equipped with a fisheye lens camera is proposed and installed at Prince of Songkla University (PSU), Phuket. The station retrieves cloud images in conjunction with several meteorological sensors. The methodology and cloud classifier algorithm is proposed. The performance test using images captured from our cloud monitoring station shows that the cloud classification accuracy in practice is as high as 98.58%. The installed cloud monitoring system can report live cloud conditions and display them on a mobile application. Our complete system is inexpensive and suitable for local use of weather monitoring and alerting.