EDITORIAL: CORRESPONDENCE AND REGISTRATION TECHNIQUES

Abstract

The objective of image correspondence is to match areas or features that are present in two (or more) images. Approaches to the correspondence problem can broadly be classified into two categories: the “area-based ” matching and the “feature-based ” matching techniques. In the first category, the matching process is applied directly to the intensity profiles of the two images, while in the second, features are first extracted from the images and the matching process is applied to the features. Other techniques have also been used for the sake of devising more robust correspondence techniques as discussed in the paper by Eklund et al. The basic problem of image correspondence can then be described as follows: given two images, find the areas or features of each image that match with a second image. For example, given two faces taken from two different viewpoints, a good image correspondence algorithm can match up the eyes, nose, mouth, ears and other facial features that are visible in the images under consideration. If it cannot find an exact match, it should find the “best” match. Correspondence is the crux of any stereo matching algorithm. It is also an important prior stage to image registration. It can also be used to automatically perform other complex image manipulation tasks that have previously been done by hand.