Abstract: Satellite image processing plays an important role to extract useful information about earth. Satellite image of earth can be used in various applications such as detecting land cover, ship detection, forestry, geology, meteorology etc. The traditional surveying method which requires a huge amount of manpower, money and time can be reduced by applying various image processing methods on satellite image. Our project aims in using two levels of classifications of which the first level of classification includes classification of land, water bodies and urban areas using unsupervised approach. The level two classification aims to divide the land data into vegetated and non-vegetated regions using PCA method. The result of above classification would give us the percent ratio of land under vegetation and non-vegetation

Keywords: PCA, Classification system