Efficient Cubic B-spline Image Interpolation on a GPU

Application of geometric transformation to images requires an interpolation step. When applied to image rotation, the presently most e cient GPU implementation for the cubic spline image interpolation still cost about 10 times as much as linear interpolation. This implementation involves two steps: a pre lter step performs a two-pass forward-backward recursive lter, then a cubic polynomial interpolation step is implemented thanks to a cascade of linear interpolations. This paper proposes a simpler and faster implementation for the pre lter which is the most time consuming in terms of a direct convolution. The overall cost for our cubic B-spline interpolation algorithm then reduces to only twice the cost of linear interpolation.