Forecasting of cyclone using multi-temporal change detected satellite images

Clouds in the sky play an important role in predicting the weather. Satellite images provide an excellent view of clouds in the sky which can be certainly used in forecasting the weather. In Satellite images, particularly Infra Red (IR) images play a vital role in many applications. The satellite cloud images contain valuable information for weather forecasting and early prediction of different atmospheric disturbance such as typhoons, hurricanes etc. Very strong winds and torrential rainfall which makes people to become shelter less during cyclones. Therefore in forecasting weather, cyclone prediction has a major role as this is directly related to the lives and households of human being. Many sub-continents in the world have regions that are affected severely because of cyclones. In India, during Madi 2013 cyclone, nearly seven lakh people were forced to relocate, to prevent the loss of human life and their assets. So it is necessary to plan out your day accordingly so that you are not caught off guard. In the literatures, for cyclone prediction which is mainly dependent on Doppler Radars along with Historical Satellite images. Because of its limited range and semi-automatic approach, there is a need to propose a method to automate cyclone prediction, which finds the movement & direction of cyclone towards the given specific locality and relays those information to the subscribed users. Fuzzy C-means clustering is applied for segmenting each cyclone image, followed by extracting features like shape, color and texture for the same. Change Detection is the process of finding changes in multitemporal images. The features obtained from the multi-temporal cyclone IR images are compared with each other in order to determine the movement of cyclone.