<![CDATA[Extreme weather caused by Cumulonimbus (CB) clouds can threaten fishermen’s safety and aviation operations in tropical Indonesia. This study designs and implements a weather monitoring system based on Meteor M2 satellite imagery with automatic CB cloud detection using a Double V-Dipole antenna, SPF5189Z Low Noise Amplifier (LNA), RTL-SDR V4, SatDump, and the YOLOv8 model. The antenna was designed at 137.5 MHz to receive Low Rate Picture Transmission (LRPT) signals from Meteor M2-3 and M2-4 within the 137.1–137.9 MHz frequency range. Fabrication testing achieved a return loss of −25.5077 dB, a VSWR of 1.112, and an impedance of 50.474 Ω. Signal reception achieved an average peak SNR of 15.246 dB using the fabricated antenna, with all test sessions successfully producing GeoTIFF imagery. The YOLOv8 model was trained using 4,346 annotated instances and achieved an mAP@0.5 of 0.427 and a maximum F1-score of 0.46 at a confidence threshold of 0.177. The detection results and cloud-top temperature estimates were visualized through a Flask-based dashboard and an interactive Leaflet.js map. The developed system demonstrates the integration of satellite image reception, CB detection, and weather visualization using open-source technologies and low-cost devices.]]>
Copyrights © 2026
