obtained to the fusion decision map, then applied to the decision map bythe transform domain. Finally, the resulted image will have more detailsof both the source images..
Registered TransformImage co-efficients
Fig. 1. Block diagram of Transform domain image fusionThe main concept and theory of wavelet based multi-resolution isdelivered from mallrt. It can detect local features in a signal process and itsi used to decompose process. Besides texture analysis, data compressionand feature detection, it can also be used for image fusion. Even beforewavelet based transform, the pyramid base transform was introduced byBurt and Adelson in 1983 and then it was improvised by Toet .But inthese type of techniques are not supported to continuous function. So thewavelet technique is applied to the image fusion.In common , all types of transform domain fusion techniques , thetransformed image are combined in the transform domain using any of thefusion rule, it is denoted by
of the registered input images I
(x,y) and I
(x,y) together with the fusion rule φ then the inverse wavelet transform W
is applied and the fused image I (x,y) is reconstructed.
I(x,y) = w
Registered wavelet fused wavelet FusedimageInput Images coefficients coefficients ImagesFig.2. Fusion of the Wavelet Transform of two images
III DISCRETE WAVELET TRANSFORM
The mainidea of all multi-resolution schemes centers around the humanvital system being local contrast ie., edges or corners. Two registeredimages of the same scene each of the coefficients of each transform possess significantly different magnitudes within the regionUsing anyone of the fusion rule to generate the combined coefficients map then theinverse DWT is applied to the combined coefficients map to produce thefused image which is give the more information of than the input images.Using separate Filter and down-sampling in the horizontal and verticaldirections produces four subbands at each scale. It denotes the horizontalfrequency and then the vertical frequency. This produces high-high(HH),high-low(HL), los-high(LH) and low-low(LL) image subbands. Byrecursively applying the same scheme to the low-low subband a multi-resolution decomposes process can be achieved.(Fig.3)Fig.3. Labelled subbandsFusion Rules: there are three fusion rulesthatwere developed andimplemented using DWT based image fusion.
maximum selection (MS) scheme:
This simple scheme just picks the coefficient in each subband with the largestmagnitude;
weighted average (WA) scheme:
This scheme developed byBurt and Kolczynski  uses a normalised correlation between the two images’ subbands over a small local area. Theresultant coefficient for reconstruction is calculated from thismeasure via a weighted average of the two images’coefficients.
window based verification (WBV) scheme:
This schemedeveloped by Li
creates a binary decision map to choose between each pair of coefficients using a majority filter.In DWT fits most commonly into the decimated form(Mallats DyadicFilter Tree0. It is only used for compression but other occus in other signal analysis such as lack of shift in-variance. Sice the wavelet fitnessare separable, the small shift in the input signal can cause major variationsin the energy distribution between DWT coefficients at different scalesand poor directional selectivity for diagonal features. So the new approachwas introduced to shift in-variance tha is known as Complex Wavelet.However, the real valued wavelet transform suffers from shift varianceand lack of directional selectivity.
Nikolov et al.
 introduced the use of the dual tree complex wavelet transform (DT-CWT) for image fusion. TheDT-CWT is approximately shift invariant and has double the amount of directional selectivity compared to a real valued wavelet transform. Shiftinvariance is an important feature of a fusion transform as the magnitudeof the coefficients of a shift variant transform may not properly reflecttheir importance. The improved directional selectivity of the DT-CWT isalso important in order to properly reflect the content of the images across boundaries and other important directional features. In DT-CWT givesmuch better directional selectivity when multi-dimensional signalisfiltered..
IV. DUAL-TREE COMPLEX WAVELET TRANSFORM
The motivation of suing the DT-CWT for image fusion is, reduced theshift in-variance and directional selectivity when compared with theDWT. In DT-DWT comprising two trees of real filter a and b, which produce the real and imaginary parts of the complex coefficients and oddand even length bi-orthogonal linear phase filters.