Template deformation by maximizing mutual information
Mark B. Skouson, Zhi‐Pei Liang · 2003
This paper addresses the problem of fluidic template matching with application to image segmentation. The authors extend the technique by Miller et al. [see G.E. Christensen et al., IEEE Trans. Image Processing, vol. 5, p. 1435-47, 1996] to handle images with arbitrary contrast differences by using mutual information as the similarity metric.