Retrieval Techniques for Feature Extraction in CBIR System

Vanita M. Yadao, Ashish B. Kharate

Abstract: CBIR (Content-Based Image Retrieval) uses the visual contents of a picture like global features-color feature, shape feature, texture feature, and local features-spatial domain present to signify and index the image. CBIR method combines global and local features. In this paper worked on Hear Discrete Wavelet Transform (HDWT) for decaying an image into horizontal, vertical and diagonal region and Gray Level Coo-ccurrence Matrix (GLCM) for feature extraction. In this paper for classification process, Support Vector Machine (SVM) used. The experimental results show improved results in comparison to previous methods. In this paper, proposed a calculation which consolidates the advantages of a few different calculations to improve the exactness and execution of recovery.

Keywords: CBIR, Similarity Matrix, DWT, SVM, GLCM, Global feature, Local feature, Color Correlogram, Color Histogram

How to Cite?: Vanita M. Yadao, Ashish B. Kharate, "Retrieval Techniques for Feature Extraction in CBIR System", Volume 8 Issue 4, April 2019, International Journal of Science and Research (IJSR), Pages: 65-68, https://www.ijsr.net/getabstract.php?paperid=ART20196585, DOI: https://dx.doi.org/10.21275/ART20196585

Download Citation: APA | MLA | BibTeX | EndNote | RefMan