Mehran University Research Journal Of Engineering &
Technology (HEC Recognized In Category "X")
Publishing Since 1982.



For Authors
For Readers
Article Information
An Adequate Approach to Image Retrieval Based on Local Level Feature Extraction

Keywords: Content Based Image Retrieval, Feature Extraction, Feature Vector, Similarity Measure, Edge Histogram Descriptor, Fourier Descriptor

Mehran University Research Journal of Engineering & Technology

Volume 34 ,  Issue 4

Sumaira  Muhammad Hayat Khan,Ayyaz   Hussain,Imad Fakhri Taha  Al Shaikhli

Abstract

Image retrieval based on text annotation has become obsolete and is no longer interesting for scientists because of its high time complexity and low precision in results. Alternatively, increase in the amount of digital images has generated an excessive need for an accurate and efficient retrieval system. This paper proposes content based image retrieval technique at a local level incorporating all the rudimentary features. Image undergoes the segmentation process initially and each segment is then directed to the feature extraction process. The proposed technique is also based on image?s content which primarily includes texture, shape and color. Besides these three basic features, FD (Fourier Descriptors) and edge histogram descriptors are also calculated to enhance the feature extraction process by taking hold of information at the boundary. Performance of the proposed method is found to be quite adequate when compared with the results from one of the best local level CBIR (Content Based Image Retrieval) techniques.