Abstract: Preserving edge information is a difficulty in image interpolation.A novel algorithm for image enlargement through quadtree decomposition is proposed.The main idea is to partition the input image into quadrants using quadtree decomposition，quadtree decomposition would divide the original image into small blocks.When the block is 1×1，it uses nearest-
neighbor interpolation.The block is 2×2，it uses bilinear interpolation.Otherwise bicubic interpolation is used，then it maps the enlarged block into the output image.Simulation results show that the new interpolation technique can preserve edge structure and get good efficiency.The proposed method can be used for image/video superresolution applications and medical image processing.

Key words: image interpolation, edge preservation, quadtree decomposition

摘要： 如何在图像插值的同时较好地保持图像的结构信息一直是图像插值的一个难点，提出了一种基于四叉树分解的图像插值方法，基本思想是：用四叉树分解将原始图像划分为一些像素值相近的子块，当分解区域为1×1，用最邻近插值方法，分解区域为2×2时，用双线性方法插值，其余情况用双立方插值扩充子块，再将扩充后的子块映射到输出图像。实验结果表明该方法能较好地保持图像的边缘，同时具有较高的效率，可以应用到图像、视频及医学图像的放大。

关键词: 图像插值, 边缘保持, 四叉树分解