Premium
Image interpolation with adaptive receptive field‐based Gaussian radial basis functions
Author(s) -
Ahmed Farid,
Gustafson Steven C.,
Karim Mohammad A.
Publication year - 1996
Publication title -
microwave and optical technology letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.304
H-Index - 76
eISSN - 1098-2760
pISSN - 0895-2477
DOI - 10.1002/(sici)1098-2760(199611)13:4<197::aid-mop7>3.0.co;2-h
Subject(s) - interpolation (computer graphics) , subpixel rendering , radial basis function , artificial intelligence , gaussian , pixel , receptive field , gaussian function , mathematics , fidelity , artificial neural network , bilinear interpolation , computer science , computer vision , algorithm , pattern recognition (psychology) , image (mathematics) , physics , telecommunications , quantum mechanics
Image interpolation with radial basis function (RBF) neural networks is accomplished. An RBF network is first trained with an image so as to satisfy the gray‐value constraints at each pixel. Each pixel is then divided into subpixels, and the subpixel gray values are calculated with the trained network. Two‐dimensional Gaussian basis functions are used in the neurons of the hidden layer. A distortion measure of interpolation is provided. It may be used to determine an acceptable range of receptive field width. This range can then be employed in an adaptive scheme for the determination of receptive field width of the Gaussian functions that renders a high‐fidelity interpolation with the preservation of edge details. © 1996 John Wiley & Sons, Inc.