Image Enlargement Research based on New Power Series - Newton Interpolation Algorithm

Abstract

Image interpolation is one of the important techniques in image enlargement. Traditional linear interpolation and bi-cubic interpolation methods are unable to guarantee the derivative continuity at the end point of the interpolated interval. In the case of large-scale enlargement, there will be a Block or fuzzy phenomenon. In this paper, a new power series - Newton interpolation algorithm is proposed to achieve high resolution enlargement. The interpolation principle based on the new algorithm is also analyzed: At first, the programme construct the new power-series function, then this function is used to get the derivative, finally the derivative is utilized to construct the Newton-interpolation function. The test result shows that this algorithm can well reflect the gray-level change even in large-scale enlargement, can make up for the defects of the traditional algorithms.