Four Walsh Transform Sectors Feature Vectors for Image Retrieval from Image Databases 

Abstract

The work of this paper presents the new idea of using Walsh transform to generate the feature vector for content based image retrieval. The complex Walsh transform is conceived by multiplying Sal functions by j and combining them with Cal functions of the same sequency. The angle is calculated by taking tan inverse of sal/cal  in the range of 0 to 360 degrees which  is divided into 4 sectors. The mean of real and imaginary values of these sectors in all three color planes are considered to design the feature vector of 24 components all together. The algorithm proposed here is worked over database of 270 images spread over 11 different classes. The Euclidean distance is used as similarity measure. Average precision and recall is calculated for the performance evaluation. The overall average of cross over points of precision and recall is above 50%. 

Authors and Affiliations

H. B. Kekre , Dhirendra Mishra

Keywords

Related Articles

METADATA STANDARD HARVESTING

The rapid growth of Internet resources, digital collections and libraries are constructed with the help of metadata schemas. Each metadata schema has been designed based on the requirements of the particular user communi...

Unsupervised Analysis of Arrhythmias using K-means Clustering

The Electrocardiogram provides the valuable information regarding the cardiovascular diseases. Various methods for classification of arrhythmias have been developed by researchers. Classification can be supervised or uns...

Retrieving Business Applications using Open Web API’s Web Mining – Executive Dashboard Application Case Study

Web mining is new area of research in information technology; so many   business applications that utilize data mining and text  mining techniques to extract useful business information on the we...

PERFORMANCE COMPARISION OF QOS STABILITY METHODS IN WIMAX NETWORKS

Recently IEEE 802.16 WiMAX  has attracted a lot of attention in wireless networking research and applications. An attempt had been made to compare DropTail, RED, Proportional fairness and DRR protocol...

An Abstract memory model describing the interaction between thread and memory with debugger tools

This paper describe the multithreaded execution and data race detectors which are commonly viewed as debugging tools.The C++ Standard defines single-threaded program execution. Basically, multithreaded execution require...

Download PDF file
  • EP ID EP91803
  • DOI -
  • Views 106
  • Downloads 0

How To Cite

H. B. Kekre, Dhirendra Mishra (2010). Four Walsh Transform Sectors Feature Vectors for Image Retrieval from Image Databases . International Journal of Computer Science and Information Technologies, 1(2), 33-37. https://europub.co.uk/articles/-A-91803