A comparative study of Computational Intelligence based techniques in the field of remote Sensing Image classification

Remote Sensing Image classification is one of the major research areas due to its wide spectrum of applications including natural terrain feature classification, land use monitoring, ground water exploration, environmental disaster assessment and urban planning etc. All these applications give a great success to terrain use but the only thing required is the proper classification of remotely sensed image. For this, there is the need of a high level Computational Intelligence based classifier for the perfect use of land cover features. Computational Intelligence explodes its area based on Swarm Intelligence, Modelization of human mind, nature inspired and some other intelligent techniques. From these major categories, we included Artificial Bee Colony optimization, Cuckoo Search, Rough Set, Fuzzy Set, Membrane Computing, Minimum Distance Classifier and Maximum Likelihood Classifier for the comparison classification of Alwar region, India. Kappa coefficient is taken as acceptance estimation parameter. User's accuracy & Producer's accuracy are considered to check the accuracy of a solitary land feature. In this paper, we want to scrutinize that which computational Intelligence based classifiers give appropriate results under same circumstances to optimize this land features.