Access to eBooks is limited to institutions that have purchased or currently subscribe to the SPIE eBooks program. eBooks are not available via an individual subscription. SPIE books (print and digital) may be purchased individually on SPIE.Org.

Contact your librarian to recommend SPIE eBooks for your organization.
Chapter 12:
Additional Fusion Applications

There are many examples of image fusion that help demonstrate diverse applications. This chapter presents results from different medical and surveillance datasets using common image fusion methods, such as wavelets and pyramids. These results are instructive as alternative methods not previously described in the book and compared to the standard wavelet and pyramid methods.

12.1 Iterative-Wavelet-Fusion Example

Advanced wavelet-based fusion algorithms can be used for the kernel of an iterative image fusion process. The fusion process has two feedback-control parameters: the number of decomposition levels and the order of wavelets. The fusion result is measured and optimized by the IQI. The iterative-wavelet-fusion algorithm is tested with commonly used images and compared with the pyramid method and the regular DWT algorithm.

Of the many image fusion methods, the DWT and various pyramids (e.g., Laplacian pyramid) are among the most common and effective versions. For quantitative evaluation of the quality of fused imagery, the root mean square error (RMSE) is the most reasonable measure of quality if a "ground truth" image is available; otherwise, the entropy, spatial frequency, or IQI can be calculated and evaluated. Here, an aDWT method that incorporates PCA and morphological processing into a regular DWT fusion algorithm is presented. Specifically, at each scale of the wavelet-transformed images, a principle vector was derived from two input images and then applied to two of the images’ approximation coefficients. For the detail coefficients, the larger absolute values were chosen and subjected to a neighborhood morphological processing procedure that served to verify the selected pixels by using a filling and cleaning operation. Furthermore, the aDWT has two adjustable parameters: the number of DWT decomposition levels and the length of the selected wavelet that determinately affect the fusion result.

An iterative fusion process is implemented that is optimized with the IQI metric. Experimental results are shown for four types of inhomogeneous imagery. The iterative aDWT (aDWTi) achieved the best fusion compared to the pyramid or DWT methods, judged on both the IQI metric and visual inspection.

Online access to SPIE eBooks is limited to subscribing institutions.

Back to Top