Development of Wavelet Transform-Based Image Fusion Technique with Improved PSNR for CT and PET Images in Comparison with Discrete Cosine Transform-Based Image Fusion Technique

Abstract

The image fusion process is characterized as gathering all relevant data from multiple images and combining it into a smaller number of images, typically only one. The aim of image fusion is to merge data from multiple images of the same scene into a single image that, in theory, contains all of the original images' important features. As a result, the resulting fused image would be better suited to human and machine perception as well as image processing tasks. Materials and methods: In this work, wavelet transform-based image fusion technique was proposed and developed to improve the quality of images. The proposed work is compared with discrete cosine transform based image fusion technique using Peak Signal to Noise Ratio (PSNR) and Universal Image Quality Index (UIQI) values and sample size for each group is 30. The threshold value is set to 0.05 and the confidence interval as 95%. Results: The performance of image fusion of both algorithms are measured by using parameters PSNR and UIQI. High values of PSNR and UIQI indicates the better image fusion. Discrete wavelet transform provides mean value of PSNR (15.8460 dB), mean value of UIQI (0.1966%), and discrete cosine transform provides the mean value of PSNR (15.8783 dB), mean value of UIQI (0.0802%). The insignificant values of PSNR 1 (P=0.894), PSNR 2 (P=0.643), UIQI (P=0.057). Conclusion: Based on the experimental results, it is observed that the discrete wavelet transform-based image fusion method performs better discrete cosine transform.