Boundary effects in wavelet transform and improved algorithm in image fusion

Image boundary effects occur in image fusion due to wavelet transforms. To overcome it, causes leading to boundary effects are analyzed, based on the definition of wavelet and the boundary extension methods in its reconstruction. According to the analysis, the traditional boundary extension method is modified. The method introduced by shortening the signal followed by its extension satisfies the demand of perfect reconstruction. It is proved through experiments that the image fusion algorithm with the modified signal extension method eliminates the boundary effects in image fusion, and improves the quality of image fusion.