Image classification using NPR and comparison of classification results

Image classification is an important field of artificial intelligence. A machine tries to classify the images into different predefined categories using the intelligence and experience gained by the training session. Classification is a complex process which is affected by many factors as it is difficult to classify an image containing blurry and noisy content. The main aim of this paper is to provide a brief overview of different classification methods and classification techniques. In this paper three data sets are taken namely cancer, iris and wine dataset. These data sets are optimized using GA-LDA and then NPR tool is used for further training and classification process. After the classification the results are compared to make a final conclusion that as the dimensionality increases, the number of selected features should increase up to a certain number for the better classification results.