Image Classification- Based on Fuzzy C means Clustering Algorithm: Sub pixel classification using Remote sensed data

The analysis of Fuzzy Logic is carried out by varying the number of classes and also by changing the number of classes per pixel. Different classification methods yields different results and none of the methods is suitable for variable classes and variable pixels. Depending upon the information needed about the classification, suitable classes and suitable classes per pixel should be selected. The selection of a particular method is dependent on number of classes and number of classes per pixel. The suitability of a particular scheme depends to some extent on the nature of the image to be classified. When results of all the classified methods considered were compared, application of Fuzzy Logic only makes the classification more complex. It is because of the type of the data used for classification. Since the data used is a 23.5m resolution data, which is considered as low resolution data, almost all the methods produce reasonably high accuracy value. It is hard to come to a conclusion by visually examining the classified images. Hence an accuracy assessment is carried out for numerically/quantitatively finding the classification method that results in highest accuracy value.