Authors : Raghoba Nagpure and Sanjay Balamwar
Page Nos : 33-43
Description :
With a view of accomplishing the needs for further analysis on image, a remotely sensed image is at first pre-processed in order to remove anomalies from it, thus resulting in clear and informative image. The importance of the accuracy in processing of a remotely sensed image cannot be over emphasized as without it the quality of output produced would be of lesser value to the end user. Remote sensing image classification involves supervised and unsupervised techniques. Both the techniques give different outputs and levels of accuracy. This paper describes the analysis that was carried out to perform supervised and unsupervised techniques on remote sensing data for land cover classification and to evaluate greenness in terms of area over a period of time. Both the methods are used for object classification and detection. The input images were enhanced using histogram equalization technique and then segmented using supervised and unsupervised classification with the help of ERDAS software which equips user with a dynamic set of tools for executing numerous task like geo correction and its analysis, data visualization, and mapping of outputs. After segmentation or image classification results were analyzed to give exact measure of greenness in terms of area in hectares and square kilometers calculated over a span of years.
