AbstractAn diagnosis, ocean surveillance etc. For fusing the

AbstractAn image fusion is uniting two or more images of common characteristics to form a single image which has access to all necessary features of the original image. In present day era a lot is happening in the field of image fusion with various applications such as battlefield intelligence, medical diagnosis, ocean surveillance etc. For fusing the image numerous techniques had been proposed by several authors such as High pass filtering technique, Wavelet transform based image fusion, IHS transform based image fusion, PCA based image fusion etc.                                 Here in this paper we discuss about the literature of image fusion with wavelet transform along with the merits and demerits of itself.INTRODUCTIONImage fusion is combination of two images resulting into a more relevant image without producing details that are not present in the given images. With hasty advancements in the field of technology it is now possible to obtain a high quality fused image with spatial and spectral information too. Image fusion is the process to improve the set of information from given set of images. Image fusion actually helps in increasing reliability and decreasing uncertainty in the image.  Image fusion can be performed at the three different levels: pixel, feature and decision level.          Image fusion is basically divided into categories: Direct Image fusion and Multi resolution Image fusion. Multi resolution Image fusion is performed on pixel level and in general uses the Wavelet and Multi Wavelet transform for the representation of the input image. Wavelet transform produces much better result than the fusion based transform methods. DWT in multi resolution can produce good results in frequency and space domain. DWT provides more directional information in the LL, LH, HL, HH bands and contains exclusive directional information.DISCRETE WAVELET TRANSFORM  It is used to convert the input image from spatial image to frequency domain. For the first order of the process to be accomplished the image is divided into vertical and horizontal lines, this divides the image into 4 parts given as LL1, LH1, HL1 and HH1.          The four parts represent four different frequency areas for the given input image. LH1, HL1, HH1 have comparatively more detail than LL1. LL1 is actually not visible to human eyes.                               Fig. First order DWT PROCESS FLOW OF DWTWith the help of a set of fusion rules, the DWT is performed on each input image. After combining the coefficients obtained from the input image the inverse DWT is applied on the fused image with coefficients.                 Fig. Process flow of DWTDWT DECOMPOSITIONHigh-pass and Low-pass are the two filters present in the DWT decomposition. Other than these filters a sub-sampling process is also required for successful decomposition of an input image. The four frequency bands have their own significance, as LL band contains the approximation coefficients, LH image contains the horizontal detail coefficient while HL contains the vertical one, whereas the HH has all the diagonal coefficients. For further the decomposition only the LL image.             Fig. Decomposition in DWT