Multidimensional Reconstruction by Set-valued Approximations
David Levin · IMA Journal of Numerical Analysis · 1986
Some generalizations of the notion of univariate data interpolation are presented, inducing the concept of set-valued interpolation in a general metric space. Consequently methods for univariate interpolation or smoothing of multidimensional geometrical data are suggested. In particular the application of these methods to 3-D body recognition from cross-sectional data is discussed. Preliminary analysis of the interpolation process is presented and the capability of reconstructing bodies of complex topologies is exemplified.