A Sub-block Based Image Retrieval Using Modified Integrated Region Matching

TL;DR

This paper presents a CBIR system that combines local colour and texture features from image sub-blocks with global shape features, using a modified IRM algorithm to improve image retrieval accuracy.

Contribution

It introduces a novel sub-block segmentation method and a modified IRM algorithm for more effective image retrieval in CBIR systems.

Findings

Outperforms existing methods in retrieval accuracy

Effective segmentation of image sub-blocks based on edge density

Improved matching using modified IRM algorithm

Abstract

This paper proposes a content based image retrieval (CBIR) system using the local colour and texture features of selected image sub-blocks and global colour and shape features of the image. The image sub-blocks are roughly identified by segmenting the image into partitions of different configuration, finding the edge density in each partition using edge thresholding followed by morphological dilation. The colour and texture features of the identified regions are computed from the histograms of the quantized HSV colour space and Gray Level Co- occurrence Matrix (GLCM) respectively. The colour and texture feature vectors is computed for each region. The shape features are computed from the Edge Histogram Descriptor (EHD). A modified Integrated Region Matching (IRM) algorithm is used for finding the minimum distance between the sub-blocks of the query and target image. Experimental results show that the proposed method provides better retrieving result than retrieval using some of the existing methods.