Content-Based Image Retrieval: Feature Extraction Techniques and Similarity Metrics

Abstract

The method which is used here for the purpose of feature extraction keep visible qualities of the picture in mind. In this paper, various feature extraction techniques based on color, shape, texture etc. are discussed. In addition to the color bar chart and its correlogram, CCV, CCM and tamura, wavelet transform, DWT, GLCM, DCD, SCD, CSD etc. comes under feature extraction methods. Firstly, feature extraction procedure is discussed in the paper then various techniques for feature extraction is discussed. After that an algorithm is proposed to retrieve images depending on various features. Then results of different similarity metrics are shown.