Efficient Algorithms Of Image Fusion Based On Hybrid Techniques

Image fusion has number of applications such as images taken for medical-checkup,robotics,computer vision,military and law enforcement procedures.Main purpose of image fusion is to extract all useful information from input images and provide one output mage of high quality.In spite of its multiple applications,existing image fusion techniques have some significant limitations i.e.,spatial domain techniques have low frequency information where as spectral domain have low time resolution which degrades the quality of output image.However existing techniques like stationary wavelet transform(SWT)provide adequate frequency resolution but has poor time information while in discrete wavelet transform(DWT)image is transformed from spatial domain to frequency domain.In case of principal component analysis(PCA),it provides good time information but has poor frequency information.Keeping in mind the existing complications,there is a dire need to develop an efficient algorithm which can provide both frequency resolution as well as time resolution simultaneously.To compare the results of two proposed methods,PCA+DWT and PCA+DSWT;a variety of image fusion quality parameters such as peak signal to noise ratio,mean squared error and normalized cross correlation are employed for the qualitative measurements.Analysis of comparison results demonstrates that the proposed methods PCA+DWT and PCA+DSWT exhibit better outcomes than many other fusion methods,qualitatively as well as quantitatively.These hybrid techniques are proposed to obtain a better efficient and better quality fused image which will preserve edges,its spatial resolution and improve shift invariance.MATLAB tool is also used for all simulations and were carried out through whole research work.