Multilevel feature extraction and X-ray image classification
Document Type
Article
Publication Date
1-1-2007
Abstract
The need of content-based image retrieval tools increases with the enormous growth of digital medical image database. Classification of images is an important step of content-based image retrieval (CBIR). In this study, we propose a new image classification method by using multi-level image features and state-of-the-art machine learning method, Support Vector Machine (SVM). Most of the previous work in medical image classification deals with combining different global features, or local level features are used independently. We extracted three levels of features global, local and pixel and combine them together in one big feature vector. Our combined feature vector achieved a recognition rate of 89. Large dimensional feature vector is reduced by Principal Component Analysis (PCA). Performance of two classifiers K-Nearest Neighbor (K-NN) and Support Vector Machine (SVM) are also observed. Experiments are performed to verify that the proposed method improves the quality of image classification.
Keywords
image processing, Image classification, machine learning
Divisions
ai
Publication Title
Applied Sciences
Volume
7
Issue
8