A class of image interpolation method based on quasi Hermite interpolation spline

The efficiency of any image resizing algorithm is determined by two main factors: the quality of the obtained image and the computational complexity. Although many image interpolation methods have been proposed in the literatures, these methods suffer from the computational complexity as well as runtime evaluations. This paper presents a method for image zooming using quasi Hermite interpolation spline, which provides a smaller interpolation error, and reasonable computing times. Moreover, the method can preserve the contours or edges of images by adjusting the corresponding local parameters. It is a more accurate and fast approximation for the ideal interpolation in space domain than by other linear polynomial interpolation kernel functions. Simulation results are also presented to demonstrate the effectiveness of the proposed method.