Medical Image Zooming Algorithm Based on Bivariate Rational Interpolation

A new method for image processing, especially for image zooming by bivariate rational interpolation based on function values is presented. Suppose that the given data points of an image are sampled from an original surface, then the continuous interpolation surface is constructed based on rational spline to fit the original surface. The image resized can be obtained just by resampling on the interpolation surface. The new method produces pictures with higher precision, which is useful in 2D CT image resizing. According to the special property of CT image, the value of new resampling data point can be taken from a region. The experimentations show that the proposed method can maintain clear border of zoomed image and hence offer more detail information in real application for doctors. The algorithm is simple and efficient in computation.