Understanding Image Registration

In Understanding Image Registration, two applications of image registration are discussed: the retinal image registration method and the far infrared image registration method. Many applications such as aerial photography and other wide-field-of-view image registering techniques can benefit from said methods. Following this, in order to compare the performance of image registration using sparse coding with other non-sparse registration methods, a simulated transformation is utilized. Even in the presence of bias field distortion, the proposed method aligns images without any preprocessing for bias field correction. The concluding study focuses on how the concept of a Gelfand pair from the theory of noncommutative harmonic analysis points to a two-step procedure. It introduces the Mendez transform that finds a partial solution to a registration problem, and the remaining problem can then be solved more easily using well-known methods.