112. IMAGE INTERPOLATION USING CLASSIFICATION AND STITCHING
Name: Nickolaus John Mueller
Grad Year: 2012
Image interpolation is a well studied signal processing application that continues to receive substantial attention from the research community. The goal of image interpolation is to produce a higher resolution version of a given image. Basic techniques, such as bilinear and bicubic interpolation, accomplish this at a low cost, but they sacrifice quality of the final image by assuming that the underlying ?true image? is piecewise polynomial. It has been recognized that taking into account the presence of edges in an image can significantly improve the resulting interpolated image. Many papers have been proposed that modify the interpolation method in the presence of edges to avoid common artifacts such as blurring, blocking and ringing. We propose that a single interpolation method is inadequate for producing the highest possible quality of resolution enhancement. Instead of using a single method for interpolation, we fuse thecbest features of several interpolation methods. This can be done using a novel region classification algorithm to determine which method is best suited to a particular region. By segmenting the texture regions from the rest of the image, we are able to use one method to interpolate texture regions and an edge-sensitive method to interpolate the rest of the image. From this information, we can use image mosaic techniques to seamlessly fuse the methods into a result that contains both sharp edges and detailed textures.