DESIGN AND IMPLEMENTATION OF MULTI LABEL LAND COVER CHANGE PREDICTION MODEL BY MODIFIED ANN-CNN AND CA-MARKOV METHOD

The land cover map classification, change detection and prediction are essential tasks for planning and monitoring the resources on the earth for sustainability development. Many numbers of machine learning approach have been employed in the current research to bring out the relatively good model for performing the classification, change analysis and growth prediction and the ample amount effort on them were not yet used in this direction. In this paper, the modified water wave optimization with ANN-CNN algorithm has been employed to produce the better classified images for developing future land cover predictions. The multi label land cover change prediction approach has been introduced which evaluates suitability index for the land cover type. CA-Markov has been then employed to generate the stationarity distribution of the land cover types which clearly shows the demand prediction of the land cover type in the remote sensing image. At the end, the results of this models was validated and compared with other supervised classifiers. The modified ANN-CNN-CA-Markov model for classified remote sensing images generated essential and valuable information related to the change prediction of land cover type for planning the resources and monitoring the activities on the Earth's surface.