The purpose of this paper is to classify the LISS-III satellite images into different classes as agriculture, urban and water body. Here pixel based classification is used to classify each pixel of the satellite image as belonging to one of those three classes. To perform this classification, a neural network back propagation technique is used. The neural network consists of three layers: Input layer, hidden layer and output layer. During training of a network, the sample inputs are given to the input layer which then propagate the hidden layer and then later to the output layer. Each neuron in hidden layer will receive the inputs from all the neurons of the corresponding synoptic weights and summed up. Each neuron of output layer will also get the input from all the neurons of hidden layer which are also multiplied with their corresponding weights. The outputs of the output layer are compared with desired result. The error between desired output and actual output is calculated to obtain error matrix. For each input, both local and global classification is obtained. In local classification, the neural network is trained using a particular image and that same image is given as input. In global classification, a new input is given for the network rather than the images with which it has been trained. Accuracy is then calculated for both the local and global classification.

1. Benedikson, J.A., 1990. Neural network approaches verses Statistical methods in Classification of Multi source Remote Sensing data. IEEE Transactions on Geo Science and Remote Sensing .
2. Bischof . H.,1992. Multispectral classification of Landsat – images using Neural Network.IEEE Transactions on Geo Science and Remote Sensing. 30(3), 482-490.
3. Chen.Y.Q.,1995.Novel Techniques for image Texture classification. PHD Thesis, university of Southampton.
4. Coifman. R.R..1992 Entropy-based algorithms for best basis selection. IEEE Trans. On Inf. Theory, 38(2),713-718.
5. Hagan M.T. and Menhaj. M., 1992 .Training feed forward networks with the Maquardt algorithm. IEEE Trans, on Neural Networks, 5(6),989-993.
6. Heerman.P.D.and khazenie., 1992.Classification of multi spectral remote sensing data using a back propagation neural network. IEEE trans. Geosci. Remote Sensing, 30(1),81-88.
7. Hepner.G.F.,1990.Artificial Neural Network classification using mininmal training set:comparision to conventional supervised classification. Photogrammetric Engineering and Remote Sensing, 56,469-473.
8. Mohanty.k.k.and Majumbar. T.J.,1996. An Artificial Neural Network (ANN) based software package for classification of remotely sensed data. Computers and Geosciences ,81-87.
9. Rui. Xu. and Donald. Wuncsch.,2005.Survey of Clustering Algorithm. IEEE, Vol.16,No.3,pp.645-678.
10. Shiaofen Fang.,2009.Facial Image Classification of Mouse Embryos for the Animal Model study of Fetal Alcohol Syndrome. Proceedings of the ACM symposium, 852-856.
11. Wonil.Kim.,2005.Multi Class Adult Image Classification Using Neural Networks. Artificial Intelligence, LNAI 3501, 222 – 226.
12. Xiao. Yang. tan., 2005 Recognizing partially occluded expression variant faces single training image per person with SOM &soft Neural Network Enemble. IEEE Trans. Pattern Anal., 16(4),875-855.
13. Zhang .Y. 2005. Color clustering & learning for image segmentation based on Neural Network. IEEE, 16(4), 925-936.

Image classification Neural Networks back propagation algorithm LISS-III Multispectral images