Landcover Classification of VHRS Imagery Using Support Vector Machine

Land use/cover maps are in high demand for monitoring and managing natural resources, growth policies, and global change studies. Land use/cover maps are essential documents that provide data for a variety of purposes, Land use strategy creation, environment administration, and metropolitan arranging. The classification of land cover surface into different land use and land cover classes is important as it helps administrative bodies in managing transportation, healthcare, agriculture, education, and other activities. The aim of this study is to classify the very high-resolution satellite (VHRS) imagery obtained from different satellites using a support vector machine (SVM) and evaluate the performance of the proposed method on various evaluation metrics such as producer accuracy (PC), user's accuracy (UC), accuracy (AC) and kappa coefficients (KC). The proposed method has obtained an overall accuracy of 98.68 and a KS of 0.9858.