Search

Closed-form solutions for lowpass filtering of 3D digital geometry are presented. Conventional signal processing is difficult to apply since 3D digital geometry does not usually satisfy the prerequisites of regular sampling and global parameterisation. Representing digital geometry as a level set of a 3D scalar field fitted with Gaussians and polynomials, derived are closed-form solutions of convolution with Gaussian smoothing kernels and lowpass filtering without the pitfalls of other approaches are analytically achieved.

Convolutional radial basis function (RBF) networks are introduced for smoothing out irregularly sampled signals. The proposed technique involves training an RBF network and then convolving it with a Gaussian smoothing kernel in an analytical manner. Since the convolution results in an analytic form, the computation necessary for numerical convolution is avoided. Convolutional RBF networks need training only once, do not depend on any particular details of the training methods used, and different degrees of smoothing are immediately available.