Image Classification by Combining Wavelet Transform and Neural Network

In present-day classification of multi-class image play an important role in engineering and computer vision application like image processing in biomedical, retrieval of content based image. From some past years researchers and scientists have made a lot of efforts in the implementation of an advanced image classification approaches. Image classification is one of the classical problems of concern in image processing. Image classification is both challenging as well as important task. There are various approaches for solving this problem. In this proposed method our objective is to successfully classify an image from given large image database. Image features which contained most important information for successful classification are extracted by using Haar wavelet and Daubechies wavelet (db4) wavelet discrete Mayer wavelet (demy).In this proposed method received image features are first used with ANN for training and testing and then used same image features of different wavelet transform for KNN training testing. Finally, we evaluate the performance of both ANN and KNN classifier with different wavelet Features. Highest classification efficiency is received by Dmey based ANN classifier.