Severe weather events (e.g. cyclones, thunderstorms, fog, floods etc.) have catastrophic effects on human life and aggrandize the impacts on agriculture, economy etc. Therefore information related to severe weather, in any form, is important for mitigation purposes and helps in delineating the causes and influences responsible. Satellites images provide such information related to the formation of such events at early stages and help in tracking its spatial and temporal development. This paper presents a framework to obtain maximum information of severe weather events from satellite panoramas. The backbone of this proposed framework involves high quality filtering of noise followed by a region growing technique which seeds each pixel, thus providing refined information for further extraction. This approach will be very helpful not only for acquiring valuable information of developing severe weather events, but also for other fields (e.g. medical and astronomy) where high precision imaging is required.