Analysis of Satellite Images Using Supervised Image Classification Methods
Remote sensing is the processing and comparison of evidence from an object without direct contact with the target and with on-site observation.. In many fields, remote sensing is used, including Earth Science, land survey and most geography disciplines, as well as military, protection, industrial, research, planning, and humanitarian applications. Remote sensing analysis also offers a means for satellite image tracking, allowing a vast amount of data to be analyzed for image analysis. Satellite Images encompasses broad geographic spread and outcome in the use of more data which include classification into different divisions. There are already more than one classification algorithms available for satellite image classification but an algorithm with better performance and accuracy is required for the wide range of applications. In this paper, four kinds of supervised classification such as minimum distance, parallelepiped, k-Nearest neighbour and maximum likelihood are analysed and a comparative analysis on their effectiveness is given.
Keywords- Image Classification, Supervised classifiers, K-nearest, Satellite image.