A region based SAR images classification using Improved Moment Analysis (IMA)

摘要

Image classification is one of the major research areas of mining. Classification is basically to classify the images into similar groups. The classification is performed relevant to the type of contents it contains such as the color featured classification, texture images classification etc. In this paper, region based SAR images classification is performed. The complete work is the mix approach that will use the concept of histogram analysis along with moment analysis to perform the image Classification. The work is divided into two stages, the histogram analysis will improve the efficiency and the moment based analysis will improve the accuracy of the work. Different regions are identified based on peak value analysis, while the first and the second approach will perform the colorization and region identification of the image. Finally SAR image classification is performed using Improved Moment Analysis (IMA) and obtained results are compared with other competitive classifiers result.