Improvement of Image Fusion by Integrating Wavelet Transform with Principal Component Analysis

The process which combines the two or more than two related source images and gives a single output image is known as image fusion. Image fusion is mainly used to analyze the image areas where the pixel values i.e. information is low intensity. Fusion of Images has been used in different applications .Correlation property is important in image fusion analysis. Correlation can be controlled by distributing the Energy in different spectral bands. Broadly image fusion process can be categorized into three groups i.e. spatial, transform and statistical methods. The image fusion process should preserve suitable pattern information from all source (input) images. Average method, Principal component Analysis is comes under spatial domain method, which deals with directly changing the pixel values but the spatial domain method introduces a spatial distortion for fused image. Wavelet based image fusion is a transform domain method which gives better performance than the spatial method. We presented a novel fusion technique which is implemented by integrating the wavelet transform with Principal Component Analysis and compared the performance with respect to different performance metrics.