Image fusion techniques for high resolution images — A comparison

Fusing of images is the process by which two or more images are combined into a single image by retaining the vital features alone from each image. Image fusion is a popular technique and has wide range of applications in various domains like remote sensing, robotics and medical applications. In the past years, various image fusion techniques were developed. This study focuses on twelve techniques namely Higher Order Singular Value Decomposition, Multi-resolution Singular Value Decomposition, Laplacian Pyramid Based, Gradient Pyramid Based, Principal Component Analysis, Averaging, Discrete Cosine Transforms, Discrete Wavelet Transforms and Neural Network Based Fusions. Each fusion technique is best suited for different applications based on the requirement. The fusion methods are analysed with the help of quantitative measures and the results are discussed.