A method, system, and product for separating an image into a set of layers (background, foreground, and selector) and processing those layers to produce an output image of reduced size with improved compression and greater image quality. Each pixel of image data is processed based on signals produced from the set of layers and their attributes. The method includes determining values for signals being processed based on a look-up table or template, as well as changing values based on each pixel and its neighboring pixels.