IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

A Fast New Rotation Insensitive WP-Based Method for Image Indexing and Retrieval

A Fast New Rotation Insensitive WP-Based Method for Image Indexing and Retrieval
View Sample PDF
Author(s): Saif alZahir (The University of North British Columbia, Canada)
Copyright: 2018
Pages: 14
Source title: Computer Vision: Concepts, Methodologies, Tools, and Applications
Source Author(s)/Editor(s): Information Resources Management Association (USA)
DOI: 10.4018/978-1-5225-5204-8.ch082

Purchase

View A Fast New Rotation Insensitive WP-Based Method for Image Indexing and Retrieval on the publisher's website for pricing and purchasing information.

Abstract

Large multimedia databases and digital image archival systems are being created in government, academia, military, hospitals, digital libraries, and businesses. Efficient methods to retrieve images from such large databases have become indispensable. In this chapter, the authors present a novel Wavelet Packet (WP)-based method for image identification and retrieval that enables the recovery of the original image from a database even if the image has been subjected to geometric transformations such as size-conserving rotation or flipping operations. The proposed method uses the correlation of wavelet packet coefficients to create an image signature. This signature is comprised of two parts. The first part is a short signature, SS, that represents the location of specific values of the WP coefficient correlations in each frequency band. The second portion is the basis signature of the image, which is a long signature, LS, of 1296 correlation points produced by summing up the correlation values along all frequency bands. Computer simulation results show that the method is extremely fast, has a perfect image retrieval rates (100%), and perfect geometric transformations recognition, if any. In addition, the simulation results show that target images are perfectly identified from an image database of 7500 image signatures within a short period of time (nearly 8 seconds on the average). This method is robust against geometric transformation and requires minimal data transfer and can be used for online image retrieval.

Related Content

Aswathy Ravikumar, Harini Sriraman. © 2023. 18 pages.
Ezhilarasie R., Aishwarya N., Subramani V., Umamakeswari A.. © 2023. 10 pages.
Sangeetha J.. © 2023. 13 pages.
Manivannan Doraipandian, Sriram J., Yathishan D., Palanivel S.. © 2023. 14 pages.
T. Kavitha, Malini S., Senbagavalli G.. © 2023. 36 pages.
Uma K. V., Aakash V., Deisy C.. © 2023. 23 pages.
Alageswaran Ramaiah, Arun K. S., Yathishan D., Sriram J., Palanivel S.. © 2023. 17 pages.
Body Bottom