International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 26 - Number 10 |
Year of Publication: 2011 |
Authors: Amutha A.L., Kavitha S. |
10.5120/3141-4335 |
Amutha A.L., Kavitha S. . Features based Classification of Images using Weighted Feature Support Vector Machines. International Journal of Computer Applications. 26, 10 ( July 2011), 23-29. DOI=10.5120/3141-4335
In the recent research era analyzing and classifying images into meaningful categories using low-level visual features and high-level semantic features is a challenging and important problem. This paper focuses on the classification of COREL dataset images into its specific category using Weighted Feature Support Vector Machines (WFSVM) and the results are compared with Support Vector Machine (SVM) for validation. In WFSVM, the kernel function is precomputed by assigning more weight to relevant features using the principle of maximizing deviations. Initially, any two classes of COREL dataset is divided into training and test set and segmented using Fuzzy C Means clustering. Then from each segment of the image, color and texture features are extracted. The extracted features of the training dataset are used to construct the weighted features and precomputed linear kernel for training the WFSVM and its model file is created. Using this model file the features of test samples are classified into its specific category. Overall accuracy of classification using WFSVM is 99%, and the number of support vectors created is 6 whereas the accuracy of traditional SVM is 97% and the number of support vectors created is 12, justifies the performance of the proposed method with the existing methods.