Abstract: Method and apparatus for segmenting a first region and a second region. A method for defining a boundary separating a first region and a second region of a digital image includes determining using a learning machine, based on one or more of the color arrangements, which pixels of the image satisfy criteria for classification as associated with the first region and which pixels of the image satisfy criteria for classification as associated with the second region. The digital image includes one or more color arrangements characteristic of the first region and one or more color arrangements characteristic of the second region. The method includes identifying pixels of the image that are determined not to satisfy the criteria for classification as being associated either with the first region or the second region. The method includes decontaminating the identified pixels to define a boundary between the first and second regions.

Abstract: A computer-based method of processing a computer graphics illustration having pieces of artwork includes mapping outlines of the pieces of artwork onto a grid of cells, determining a number of outlines that map to a cell, and identifying a complex region based on the determined number of outlines that map to the cell.

Abstract: A computer-based method of processing a computer graphics illustration having pieces of artwork includes mapping outlines of the pieces of artwork onto a grid of cells, determining a number of outlines that map to a cell, and identifying a complex region based on the determined number of outlines that map to the cell.

Abstract: Methods and apparatus, including computer program products, for flattening an image. In one aspect, the invention provides a method that includes receiving an image having multiple objects with each object having an original outline. Objects from among the multiple objects are selected to be abstracted objects. A planar-mapping outline is defined for each object. The planar-mapping outline of each selected object is an enclosing outline that encloses the object, and the planar-mapping outline of each object not selected is the original outline of the object. The image is divided into non-overlapping regions by using the planar-mapping outlines of all the objects. Each region is flattened.

Abstract: Methods and computer programs for processing a computer graphics illustration having overlapping pieces of artwork. The method includes flattening at least some of the pieces of artwork into non-overlapping regions, storing information describing how the appearances of one or more pieces of artwork can be combined to determine an appearance for at least one of the non-overlapping regions, determining an appearance for at least one of the non-overlapping regions based on the stored information, and outputting the determined appearance.

Abstract: Methods and apparatus for blending graphics data that includes a plurality of objects. In one aspect, objects are grouped into two layers; objects associated with a first layer are drawn; objects associated with a second layer are drawn; and data drawn into the first and second layers is blended. A layer stack may be used to define a hierarchy for blending layers of the image. The apparatus may be an output display device and more particularly a printing device. The object data may be transformed to an intermediate form and the intermediate form data for the first and second layer may be blended. The apparatus may include an interpreter for receiving object data associated with two layers of an image and for receiving a layer operator for designating which objects belong to which layer; a drawing engine for drawing objects into a first and a second buffer for a first layer and a second layer, respectively; and a blend processor for blending data in the first and second buffers.

Abstract: Methods and apparatus for simulating diffusion on an input raster. The method includes receiving an input raster for a region, the input raster including one or more objects having fixed parameter values and one or more areas of the region over which the parameter is to be diffused. The method defines a first and second raster that cover the region where the first raster includes parameter values and the second raster includes set values designating locations corresponding to one or more objects. A procedure is applied to the first and second rasters to compute diffusion values of the parameter for one or more areas of the region. The procedure includes reducing the first and second rasters until a predefined end condition is met, and expanding the reduced first and second rasters back to an original size.

Abstract: A method for defining a color gradient across a region in a display space that includes defining a bi-cubic patch enclosing the region. The bi-cubic patch includes four corners having a color associated therewith and four curves connecting pairs of the corners forming a closed loop. The four curves define the color gradient across the bi-cubic patch as a function of the four colors associated with the four corners of the bi-cubic patch. The method includes defining a unit square in parameter space including a mapping function from the unit square to the bi-cubic patch such that the corners of the bi-cubic patch map to the corners of the unit square. The color of every point in the unit square is a bi-cubic interpolation of the four colors associated with the four corners of the bi-cubic patch. The method includes clipping the bi-cubic patch for display of the region.

Abstract: A raster image processor receives a file of instructions and images in any convenient page description language and blends the images in accordance with the specified blending mode and, possibly, opacity information.

Abstract: A method and apparatus for rendering a two dimensional (2D) patch associated with an object to achieve a smooth shading across the object where the 2D patch defines a complex gradient including at least one foldover edge. The method includes subdividing the patch until each subpatch includes no more than one foldover edge, grouping the subpatches into regions divided by a foldover edge, defining a ordering for displaying the regions and rendering the regions in order.

Abstract: A system for blending graphical objects using planar map representations is described. A page description representation of an input document's graphical objects is converted, at least in part, into a planar map representation. A color is assigned to each region in the planar map as a function of the region's associated paths' transparency characteristics. In one embodiment, if the graphical processing system exhausts its memory during processing, those paths already converted to planar maps are rasterized. Any subsequent processing of graphical objects is thereafter performed in raster format. In another embodiment, an output page is divided into tiles where each tile may be rasterized independent of the other tiles. In yet another embodiment, graphical objects are processed in reverse order compared to their arrangement in the page description language representation.

Abstract: A method of generating a threshold array made up of a plurality of dot profiles, each of which is made up of a plurality of black or white pixels, certain of the pixels in certain of the dot profiles being constrained to be either black or white.

Abstract: A method of equalizing the shape and size of halftone dots in a supertile having digital cells, corresponding ideal cells and pixels to faithfully reproduce an image having predetermined gray levels. Various techniques include: (1) adding offset vectors to the coordinates of pixels in digital cells; (2) determining twice as many smaller digital cells as original digital cells; and (3) determining a model digital cell and ranking pixels in remaining digital cells to mimic the ranking of pixels in the model cell.

Abstract: The invention provides an improved method for producing halftone images with digital halftone cell dimension and orientation to match a desired ideal halftone cell in an ideal halftone grid by choosing a reference ideal halftone cell size having a predetermined number of pixels and comparing the number of pixels in the newly selected halftone cells to the reference, thereby making the number of pixels in all ideal cells to be equal. Methods are provided for controlling the assignment of pixels on the boundary line between ideal halftone cells while avoiding pixel clustering.

Abstract: An improved method for displaying a character on a raster device at relatively low resolution by identifying regions of the character that improperly touch other regions of the character, and then modifying the display of that character to move or delete pixels which decrease legibility of the character, by (1) enumerating the pixels in an order determined by the path topology, (2) searching for sequences of pixels corresponding to a pointed feature in the character that undesirably touches other parts of the character, and (3) editing the bitmap to remove such contacts.

Abstract: An improved method is described for producing halftone images with halftone cell dimensions and orientation to match a desired screen grid by choosing reference halftone gray values for each pixel in a supertile of representative halftone cells, testing source pixel gray values against corresponding reference gray values and selectively displaying corresponding pixels, where the supertile includes more than one halftone cell, is composed of whole pixels, can tile the source image plane when placed side-by-side with copies of itself, and includes halftone cells on adjacent supertiles that, in tiling the plane, align exactly or closely to approximate a smooth halftone screen. The invention also includes ways of manipulating the supertile for faster output of a halftone image.

Abstract: Method for displaying characters with improved legibility at relatively low resolution by modifying the center point fill method and displaying additional pixels as needed to maintain connectivity and avoid dropout problems. Additional pixels are turned on as needed where a black section of the character crosses the midline between two adjacent horizontal or vertical pixel centers. If the intersections of a black section of a character with a midline is entirely within one pixel which is not otherwise turned on, then that pixel is turned on; if a black section of a character crosses between two such adjacent pixel reference points in parts of both pixels, and if neither pixel is already on, the pixel having more of the black section, measured along the line between pixel centers, (or other reference points), is turned on.

Abstract: A method for displaying characters on a screen or printer, particularly kanji characters. The structure of the character is represented by stems and counters, both vertical and horizontal, the counters being the spaces between stems. The character is then adjusted to be displayed. Either the horizontal or vertical counters are grouped into a first chain of counters. The non-integral counter widths of this chain are adjusted in relation to the other counter widths within the chain, selected counter widths being made equal to others within the chain in order to most faithfuly reproduce the desired character. The remaining horizontal or vertical counters, if any, are also grouped into one or more series of chains which are also adjusted, chain by chain. Then the other set of counters are adjusted in the same way.