Paper Details: Downloads: 435
Serial Number: P1120829001
Title: Neural Network Approach for Image Feature Space Classification Employing Back-Propagation Algorithm
Authors: Y.Chakrapani, K.Soundera Rajan
Abstract: This paper presents a scheme for classification of Image feature space using Neural networks. Feature extraction reduces the dimensionality of the problem and enables the neural network to be trained on an image separate from the test image. Lowering the dimensionality of the problem reduces the computations required during the search. The main advantage of neural network is that it can adapt itself from the training data. The network adapts itself according to the distribution of feature space observed during training. A gray level image is considered for this work. The standard deviation and skewness are computed for this gray level image and the domains are classified accordingly. A feature space classification for the gray level image is carried out with neural network and the network is also tested with the test data. Computer simulations reveal that the network has been properly trained and classifies the domains correctly with minimum deviation.
Keywords: Keywords: Back-Propagation algorithm, standard deviation, skewness, domain pool, neural networks, clustering
Journal/Conference: International Journal of Artificial Intelligence and Machine Learning
Volume: 8
Issue: 3
Submission Date: 7/18/2008 12:00:00 AM
Review Date: 7/25/2008 12:00:00 AM
Publishing Date: 10/19/2008 12:00:00 AM
Article Downloads: 435
Download:

Facebook