Temporal Image Registration using GCPs and Bivariate Polynomial Estimation

J.N. Sarvaiya, S.S. Patnaik, and S.Bombaywala (India)

Keywords

Ground Control Points (GCPs), Reference image, Image registration, Sensed image.

Abstract

Image registration is the process of spatially aligning two or more images of a scene taken at different times or with different sensors or from different viewpoints. The accuracy of spatial image registration directly depends on the selection of ground control points between the images that are to be registered. The image which is used as the base image – the image which is taken under ideal circumstances and which is used to compare all the distortions in the other image is referred to as the reference image. The image which has some distortions and needs to be registered to the reference image is referred to as sensed image. This paper describes an image registration technique which will find a ground control points (GCPs) in the sensed image based on the control point defined in the reference image. Finally, registration of the images is done using the least square surface fitting and bivariate polynomial of second degree. By the use of bivariate polynomial of second degree, faithful registration is obtained by the use of only a minimum of six pairs of valid GCPs. The proposed technique has been applied to images and the results obtained show the efficiency and the accuracy of the technique.

Important Links:



Go Back