The automatic generation of translation software for graphic objects

The authors discuss their effort to extend their translation architecture for textual, tabular, and mathematical material to include graphic objects. They begin with a brief overview of their testbed, the Chameleon Project. They analyze the feasibility of automatically generating translators in the graphics domain. They use an intermediate-form model for translation and establish criteria for the representation of the intermediate and variant forms if automatic generation is to be possible.<
>