Multimodal Biometric Image Fusion Using Discrete Wavelet Transform Method

Abstract

Background: Security is a key component to every application in the modern world. One of the most effective methods for ensuring security in many sectors is the use of personal characteristics.

Objectives: The objective of this study is to enhance the accuracy and robustness of biometric recognition through the utilization of a Discrete Wavelet Transform-based multimodal image fusion approach.
Methods: Multimodal biometric techniques were designed to resolve unimodal biometrics' deficiencies. In this research, discrete wavelet transform (DWT)-based multimodal fusion algorithms are proposed, in which database images are fused. Face, iris images and fingerprint are fused and the performance metric of different decomposition level were determined.
Result: This work provides a comparison of various decomposition filters for face, iris and fingerprint fused images. The result of metric performance on the fused images showed that the fused of face-iris have the highest value of entropy and principal component analysis which imply quality information to the source image.
Conclusion: The per for mance of DWT fusion techniques show that it provides better spatial and spectral localization of image information