Digital image processing by hardware using cubic convolution interpolation

TL;DR: In this article, the authors present hardware for providing pixel data by interpolation, where previously memorized weight factors corresponding to a particular site are retrieved under the influence of the outputs of addressing circuits, multiplied with the original pixel data corresponding to the site prior to convolution and the products added together to derive the new pixel data which is subsequently stored in a memory.

Abstract: Disclosed is hardware for providing pixel data by interpolation. In the hardware previously memorized weight factors corresponding to the particular site are retrieved under the influence of the outputs of addressing circuits, multiplied with the original pixel data corresponding to the site prior to convolution and the products added together to derive the new pixel data which is subsequently stored in a memory. Also disclosed is hardware for rapidly enlarging an image. In this hardware, four lines, for example, of original pixel data are read out and, according to the degree of enlargement required, a predetermined sequence of weight factors are simultaneously applied to sixteen original pixel data (four from each line) to determine the data for the site or sites in question.