| dc.description.abstract |
In recent years image classification has emerged as the most significant area of research in the field of remote sensing. Image classification helps us to acquire the geo-spatial information from the satellite data which can be useful to industries like defence, intelligence, natural resources etc. There exist various techniques like Biogeography Based Optimization (BBO), Ant Colony Optimization (ACO) etc for image classification. Here, we are applying a metaheuristic approach called Cuckoo Search in the area of image classification. The main advantage of this algorithm over other metaheuristic approach is that its search space is extensive in nature. The proposed methodology is applied to the Alwar region of Rajasthan. The image used is a 7 band image of 472 X 546 dimensions from Indian Remote Sensing Satellite Resiurcesat. This algorithm has captured almost all the terrain features and showed high degree of efficiency for almost all the regions (water, vegetation, urban, rocky, and barren) with a Kappa coefficient of 0.9465. |
en_US |