Analysis and Comparison of Different Wavelet Transform Methods Using Benchmarks for Image Fusion

TL;DR

This paper compares five wavelet transform methods for medical image fusion, evaluating their performance using various benchmarks to determine the most effective approach for combining multimodal images like CT and MRI.

Contribution

It introduces a wavelet transform-based image fusion methodology and systematically compares five different wavelet transforms using multiple quantitative benchmarks.

Findings

Wavelet transform methods differ significantly in fusion quality.

Dual-tree CWT and Q-shift dual-tree CWT outperform others in benchmarks.

The study provides insights into the suitability of each transform for medical image fusion.

Abstract

In recent years, many research achievements are made in the medical image fusion field. Medical Image fusion means that several of various modality image information is comprehended together to form one image to express its information. The aim of image fusion is to integrate complementary and redundant information. CT/MRI is one of the most common medical image fusion. These medical modalities give information about different diseases. Complementary information is offered by CT and MRI. CT provides the best information about denser tissue and MRI offers better information on soft tissue. There are two approaches to image fusion, namely Spatial Fusion and Transform fusion. Transform fusion uses transform for representing the source images at multi-scale. This paper presents a Wavelet Transform image fusion methodology based on the intensity magnitudes of the wavelet coefficients and compares five variations of the wavelet transform implemented separately in this fusion model. The image fusion model, using the Discrete Wavelet Transform (DWT), the Stationary Wavelet Transform (SWT), the Integer Lifting Wavelet Transform (ILFT) the dual-tree Complex Wavelet Transform (DT CWT) and dual-tree Q-shift dual-tree CWT, is applied to multi-modal images. The resulting fused images are compared visually and through benchmarks such as Entropy (E), Peak Signal to Noise Ratio, (PSNR), Root Mean Square Error (RMSE), Image Quality Index (IQI) and Standard deviation (SD) computations.