A novel content-adaptive interpolation

We propose a content-adaptive interpolation algorithm for digital images. The proposed algorithm makes use of a non-linear spatial varying filter (SVF), which is simple but efficient in enlarging images. First, edge orientation statistics in an image are exploited, and then the information is used to determine suitable weights of neighboring pixels when interpolating the current pixel. Using the proposed approach, we can generate pixel values to replace missing ones such that sharp edges can be accurately reconstructed. Simulation results show that the proposed algorithm obtains better results in both subjective and objective quality measurements over existing methods.