A New Colour-Texture Feature Extraction Method for Image Retrieval System Using Gray Level Co-occurrence Matrix

Main Article Content

Fatin Mahdi
Abdulkareem Ibadi

Abstract

Proposed a new colour-texture feature extraction method is presented for Content Based Image Retrieval (CBIR) system using Gray Level Co-occurrence Matrix (GLCM). In this method, Colour-GLCM  (C-GLCM) is extracted from each colour channel, and then computes the average of each column of GLCM matrix for each channel. In this case, we will get a feature vector include colour and texture features at the same time to achieve the objectives of any CBIR system which are; decrease the Feature Vector (FV) dimensions which consequently reduces retrieval time, and also increase the retrieval accuracy.  To perform the evaluation of the proposed CBIR system, 4000 test images have been used as query images including 500 original images were selected randomly from image database of Iraqi National Museum of Modern Art, then applying seven image transformations on each original image resulting 3500 transformations image sued as query image. The proposed C-GLCM algorithm has led to improve and increase the retrieval accuracy (93.63%) comparing with GLCM that extraction from whole gray image (87.88%) and comparing with statistical properties that extraction from GLCM feature (80%).

Downloads

Download data is not yet available.

Article Details

How to Cite
[1]
F. Mahdi and A. Ibadi, “A New Colour-Texture Feature Extraction Method for Image Retrieval System Using Gray Level Co-occurrence Matrix”, JUBPAS, vol. 27, no. 6, pp. 196-211, Mar. 2020.
Section
Articles