An efficient approach of image registration using Point Cloud datasets

Image Registration means to align two image data sets to conclude the common features and differences. This is also referred as a geometrical transformation of image data sets. In this paper we have templates that how we can minimize the time and error in surface based image registration process of 3-D data sets using Point Cloud data structure. Here we have approached the work very much in a different way than the conventional approach in which the type of data sets are varies according to the image type like if the image is 2-D then the data sets are in 2-D space else if image in 3-D then data sets in 3-D space respectively. Primarily we worked on 3-D data sets which are normally used in medical industry to store the CT images, MRI images and Tumor images and also used to make 3-D models of real objects. We have used surface based image registration method in order to register the 3-D datasets. We have worked and studied on an I.C.P algorithm which registers two 3-D data sets and find the closet points into data sets as per giving tolerance distance. We make a test on different 3-D datasets like .csv, .seabed and .xyz after getting results we conclude that .xyz data sets of point cloud data structure is far better from other 3-D data sets for image registration in context of time and in error rate.