Bilinear and cubic convolution, themost popular interpolation algorithm for digital image, are analyzed first in the paper. Based on these two algorithms, a novel self-adap tive interpolation algorithm is proposed. In this new method, when the grayscale of each interpolated point is computed, the weights of direct and indirect neighbor pixels are self-adap tively determined by the complexity of the neighborhood of the image. The experiment results show that the algorithm is superior to the traditional interpolation algorithm.