Subpixel shape smoothing based on predicted shape background information

摘要

Methods and apparatus for a predictive rendering component that may generate a rendering of a character based at least in part on predictive information regarding the background into which the character is to be rendered. Using such predictive information, the predictive rendering component may produce a rendering of a character that blends into the character background more smoothly than if the predictive background information were not used. In this way, the predictive rendering component improves upon previous implementations of font smoothing.

主权项

1. A method, comprising:
performing, by one or more computing devices:
accessing a predicted background color for a shape, wherein the shape corresponds to a bitmap generated from a mask of the shape;compositing the predicted background color with a background color of the bitmap for the shape to create a composited background;generating a rendering of the shape, wherein the rendering is based on the composited background, a color of the shape, and the mask;removing one or more elements of the composited background from the rendering of the shape; andcreating a smoothed bitmap of the shape, wherein said creating comprises compositing, after said removing, the rendering of the shape with a destination background.