Evolutionary computation applied to 3D image reconstruction

Abstract

We apply a simple algorithm for evolutionary computation to the problem of approximating a human facial 3D image by a triangular mesh under the condition that the number of data points is restricted. The essential point of this problem is to locate the data points in such a way that the resulting triangular mesh approximates the facial surface as well as possible. Selection and reproduction of the data points, based on the approximation error, are shown to be effective for solving this problem even in its simplest algorithmic implementation. We also argue that such evolutionary computation might be useful in other engineering fields.