Reproduction of Remote Sensed Images Using Artificial Neural Network

Aarsi Saini, Sukhvinder Kaur

Abstract: Remote sensed images are reproduced using Artificial Neural Network. Elements of ANN perform similar to the functions of biological neuron. Once ANN learns it never forgets. Special feature of ANN is its capability of extracting correct result even from partially corrupt input. Different topologies of the earth surface are assigned with a digital code which is used as input to ANN and the location of image taken by satellite is used as target to train the ANN. Using the algorithm the trained ANN is used to reproduce the results.

Keywords: GIS Geographic Information System, LIDARlight detection and ranging, ANN Artificial Neural Network, MSE Mean Square Error, MATLAB Matrix Laboratory, SVM Support Vector Machines