A comparison of two similarity measures in intensity-based ultrasound image registration

Abstract

We investigate two registration methods for ultrasound image relying on the intensity-based similarity measure. In the first method intensity information is given by feature points which have been extracted by using Harris corner detector. The registration similarity measure is then defined as a cost-function - error cost function. In the latter method we use the same cost-function, but the uniqueness control and region correspondence are not uniform with the first method. Given this similarity measure, parametric ultrasound image registration is stated as a minimization issue. We even exploit the polynomial technique to transform the whole image dataset and estimate the sum of square error in the first measure. The result indicates that the two methods are robust and of the order of our requirements, but the latter measure outperforms the first one.