Multi-Focus Image Fusion Method via Multiple Wavelet Bases

In this article, a new method for multi-focus image fusion via multiple wavelet bases is proposed.Firstly the wavelet transform is used to perform a multiscale decomposion on each image , then based on local gradient in the fusion to get primary fusion image .And then a set of the primary fusion images are obtained.Next these primary fusion images are fused to obtain final fusion iamge within spatial domain based on local variances weighted average rule.Experimental results show the new method is better than the traditional single wavelet base method in fusion effect.

Download Full-text

Dempster-Shafer and Multi-Focus Image Fusion using Local Distance

10.5121/csit.2021.112206 ◽

2021 ◽

Author(s):

Ias Sri Wahyuni ◽

Rachid Sabre

Keyword(s):

Image Fusion ◽

Mass Function ◽

New Method ◽

Dempster Shafer Theory ◽

Local Distance ◽

Focus Image ◽

Fusion Images ◽

Shafer Theory ◽

Quadratic Distance ◽

Local Variability

In this article, we give a new method of multi-focus fusion images based on Dempster-Shafer theory using local variability (DST-LV). Indeed, the method takes into account the variability of observations of neighbouring pixels at the point studied. At each pixel, the method exploits the quadratic distance between the value of the pixel I (x, y) of the point studied and the value of all pixels which belong to its neighbourhood. Local variability is used to determine the mass function. In this work, two classes of Dempster-Shafer theory are considered: the fuzzy part and the focused part. We show that our method gives the significant and better result by comparing it to other methods.

Download Full-text

Gradient-based multi-focus image fusion method using convolution neural network

Computers & Electrical Engineering ◽

10.1016/j.compeleceng.2021.107174 ◽

2021 ◽

Vol 92 ◽

pp. 107174

Author(s):

Yang Zhou ◽

Xiaomin Yang ◽

Rongzhu Zhang ◽

Kai Liu ◽

Marco Anisetti ◽

...

Keyword(s):

Neural Network ◽

Image Fusion ◽

Convolution Neural Network ◽

Fusion Method ◽

Gradient Based ◽

Focus Image ◽

Image Fusion Method

Download Full-text

Multi-focus image fusion method based on two stage of convolutional neural network

Signal Processing ◽

10.1016/j.sigpro.2020.107681 ◽

2020 ◽

Vol 176 ◽

pp. 107681

Author(s):

Di Gai ◽

Xuanjing Shen ◽

Haipeng Chen ◽

Pengxiang Su

Keyword(s):

Neural Network ◽

Convolutional Neural Network ◽

Image Fusion ◽

Fusion Method ◽

Two Stage ◽

Focus Image ◽

Image Fusion Method

Download Full-text

A novel multi-focus image fusion method using PCNN in nonsubsampled contourlet transform domain

Optik ◽

10.1016/j.ijleo.2015.06.019 ◽

2015 ◽

Vol 126 (20) ◽

pp. 2508-2511 ◽

Cited By ~ 25

Author(s):

Jingjing Wang ◽

Qian Li ◽

Zhenhong Jia ◽

Nikola Kasabov ◽

Jie Yang

Keyword(s):

Image Fusion ◽

Contourlet Transform ◽

Transform Domain ◽

Fusion Method ◽

Nonsubsampled Contourlet Transform ◽

Focus Image ◽

Image Fusion Method

Download Full-text

Multi-Focus Image Fusion Using Local Energy Pattern

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.145.119 ◽

2011 ◽

Vol 145 ◽

pp. 119-123

Author(s):

Ko Chin Chang

Keyword(s):

Wavelet Transform ◽

Image Fusion ◽

Discrete Wavelet Transform ◽

Discrete Wavelet ◽

Local Energy ◽

View Point ◽

Fusion Algorithm ◽

Image Capture ◽

Focus Image ◽

Better Than

For general image capture device, it is difficult to obtain an image with every object in focus. To solve the fusion issue of multiple same view point images with different focal settings, a novel image fusion algorithm based on local energy pattern (LGP) is proposed in this paper. Firstly, each focus images is decomposed using discrete wavelet transform (DWT) separately. Secondly, to calculate LGP with the corresponding pixel and its surrounding pixels, then use LGP to compute the new coefficient of the pixel from each transformed images with our proposed weighted fusing rules. The rules use different operations in low-bands coefficients and high-bands coefficients. Finally, the generated image is reconstructed from the new subband coefficients. Moreover, the reconstructed image can represent more detailed for the obtained scene. Experimental results demonstrate that our scheme performs better than the traditional discrete cosine transform (DCT) and discrete wavelet transform (DWT) method in both visual perception and quantitative analysis.

Download Full-text