In medical imaging arises the problem of matching two images of the same objects but after movements or slight deformations. We present a new method for a primitive global transformation and an improvement of a recent matching strategy which makes it more robust. The strategy consists of two steps. We consider the grey level function (modulo a normalization) as a probability density function. First, we apply a density based clustering method in order to obtain a tree which classifies the points on which the grey level function is defined. Secondly, we use the identification of the hierarchical representations of the two images to guide the image matching. The general transformation invariance properties of the representations permit to extract invariant image points. But in addition, we design a new robust coarse to fine identification of the trees which applies an implicit error measure in a prediction – correction scheme using thin plate splines to interpolate the transformation function in a finer way at each step. Therefore, we will find the correspondence between invariant points even if these have locally moved. The method’s results for matching and motion analysis on real images will be discussed.
CITATION STYLE
Mattes, J., & Demongeot, J. (1999). Tree representation and implicit tree matching for a coarse to fine image matching algorithm. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1679, pp. 646–656). Springer Verlag. https://doi.org/10.1007/10704282_70
Mendeley helps you to discover research relevant for your work.