Application of a Bivariate Rational Interpolation in Image Zooming

This paper presents a new method for image zooming. The new method is different from the traditional polynomial interpolation technique. A bivariate rational interpolation is used in the new method. Using the bivariate rational interpolation is more effective than the polynomial interpolation for image zooming, especially in processing the borders of the source image. The bivariate rational interpolation function has a simple and explicit expression, thus facilitating application of image zooming. Looking upon image data as interpolation data, an interpolation surface is constructed. According to the need of zooming, resampling the gray value on the interpolation surface with a zooming scale, the image zooming can be achieved. The advantage of the new method is the simplicity as well as the high performance. Experimentations show that the new method is able to maintain the borders of source images clearly. Furthermore, the algorithm is efficient in computation for image zooming.