Progressive modes in PDE-based image compression

Algorithms based on partial differential equations (PDEs) constitute a relatively novel class of lossy image compression methods. In this paper we introduce a practically relevant extension: We demonstrate how to incorporate progressive modes into these codecs. Since the data in PDE-based codecs is only available at irregular locations, this is a challenging task. We propose two progressive modes: The first one changes the order in which the grey values are stored, while the second additionally distributes the stored information more evenly over the file. Our experiments show that the novel codecs can outperform JPEG and even JPEG 2000 for high compression ratios.