Abstract
The boundaries of many biological objects are two-dimensional curves which evolve non-rigidly with time. This paper contains a theory and experiments for estimating motion of such curves using local shape. Beginning from first principles, a symmetric and explicitly non-rigid formulation is developed. The formulation uses the topology of correspondences and its associated geometry. A numerical implementation of the resulting algorithm is also reported. The algorithm is used to estimate non-rigid motion of the endocardium in 21 MRI image sequences of normal and post-infarct dog hearts. Over 500 correspondences were computed. The measured average return error of the algorithm was less than 1 pixel.
Cite
CITATION STYLE
Tagare, H. D. (1997). Non-rigid curve correspondence for estimating heart motion. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1230, pp. 489–494). Springer Verlag. https://doi.org/10.1007/3-540-63046-5_48
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.