Analysis and Evaluation of Classifiers Using Multi Temporal Images in Land Use and Land Cover Mapping

This paper deals with the study of analysis and accuracy prediction with regard to multi-temporal satellite image classification process (MTSICP). The main objective of this paper is to assess classification accuracy of temporal images of two different seasons (dry & wet Season) using LISS III IRS ID images of H D Kote taluk. It is found that Maximum Likelihood Classifier (MLC) provides better accuracy when compared to Minimum Distance Classifier (MDC) and MaHalanobis Distance Classifier (MHDC) in all the cases. The Kappa and Overall accuracy of MLC is much better compared to MHDC and MDC.