Abstract
Registering images that have been deformed with respect to each other involves recovery of the deformation. An elastic matching algorithm developed by D. J. Burr ( IEEE Trans. Pattern Anal. Mach. Intelligence PAMI-3, No. 6, 1981, 708–713) has been extended to use contour information for registering 2D images from different modalities. One image is modeled as being deformed with respect to another goal image. Correspondences between contours in the two images are used to stretch the deformed image toward its goal. The resulting stretched image and its corresponding contours are then used to warp it further toward its goal. This process is repeated a number of times, with decreasing image stiffness. As the iterations continue the stretched image better approximates its goal image. Registration examples of deformed synthetic and clinical images are presented.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
