Converting bitmap objects to polygons

Abstract

Methods, and apparatus implementing the methods, that provide useful resolution-independent representations of the perimeter of a bitmap object. The methods find the boundary pixels in a bitmap and identify each boundary pixel edge that actually contributes to the object perimeter. The methods distinguish objects that appear to share a common edge because they share boundary pixels, when in fact the objects are separated by a pixel width. The polygon can be encoded in a digital, compact, compressible format, and can be used to define traps in a vector-based trapping environment. A finite state machine for generating the polygon is disclosed. A bitmap object is encode by calculating a sequence of direction codes based on tracing around the bitmap object along the boundary, where each direction code represents a direction from an eight-connected pixel to an adjacent eight-connected pixel. In another aspect, the invention provides a technique for trapping mixed bitmap and vector objects in a vector-based trapping process. The bitmap objects on a page are converted to perimeter polygons, the polygons are planarized with the vector objects of the page, and the resulting planar structures are trapped.