Medical image registration based on grid matching using Hausdorff Distance and Near set

Image registration is a method that resolves the discrepancy between spatial alignments of two or more images having an identical view, taken at different times, from different viewpoints and by different image sensors. It is extensively used in many image processing tasks such as medical imaging, remote sensing, military applications and so on. In this work, we propose a new automated image registration method based on virtual grid generation and coequal feature vector matching. Firstly, corners are extracted by Harris corner detector as feature points and are constructed as a collection of virtual grid from the desired corner points. The virtual grids are compared in the source and target images using coequal feature vectors. Two congruent virtual grids are taken from the source and target image. The image transformation parameters are selected from the corresponding grid points. Once feature selection is over, congruent virtual grids are established. Hausdorff Distance and Near set is used to get the correspondence and intensity variation. The experimental results show that the proposed technique is more accurate with respect to the existing state of the art research work.