Semantically image has very meaningful categories. Classifying image using the low level feature is a challenging task. So far several methods has been used for automated machine learning in semantic image classification in this paper we have proposed a new and far more efficient method for semantic image classification using normalized vectors of WFSVM(weighted feature support vector machine). For image classification, the image data usually have a large data set on number of feature dimensions. Traditional image classification algorithms based on the SVM assign normalized automated weights to these features. The relevant and non relevant features of image are separated using these normalized vectors. Using normalized vector the efficiency and training time of SVM is improved to a greater extent. In this paper we proposed an approach to use weighted normalized vectors in place of normalized vectors. The Experiment is carried out on 256-categories database and result in better. The weighted normalization of vector has two advantages than the traditional SVM: the better performance of generalization ability and less training time. © 2012 Springer-Verlag GmbH.
CITATION STYLE
Dhariwal, S., Raghuwanshi, S., & Shrivastava, S. (2012). Content based image retrieval using normalization of vector approach to SVM. In Advances in Intelligent and Soft Computing (Vol. 167 AISC, pp. 793–801). https://doi.org/10.1007/978-3-642-30111-7_76
Mendeley helps you to discover research relevant for your work.