Edge-Based Graphics Region Detection via Split Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing printing architectures face inefficiencies in processing edge-based representations, as they do not explicitly store full boundaries of non-overlapping objects, hindering subsequent processing steps and requiring cumbersome pixel-wise outline extraction methods.
Innovation Solution
A method for graphics processing that involves determining vertices and associating them with edges to define regions, using split points to form sub-edges and process regions, allowing for efficient conversion of edge-based representations to non-overlapping object-based representations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If an edge-based representation is used to resolve overlapping regions, then rasterisation efficiency is improved, but the ability to perform subsequent processing steps requiring explicit boundaries deteriorates
Solution Approach 1:
The patent segments the edge-based representation by introducing virtual edges at split points where real edges intersect or terminate. This creates a set of non-overlapping virtual edges that segment the page into distinct regions, each with explicit boundaries. The segmentation allows subsequent processing steps to identify and handle each region independently while maintaining the efficiency benefits of the original edge-based format.
2Measurement precision
If pixel-wise outline extraction is applied to determine region boundaries, then boundary accuracy is improved, but processing time increases significantly
Solution Approach 1:
The patent performs preliminary action by pre-calculating and storing split points and virtual edges during the edge-based representation generation phase. Instead of extracting boundaries pixel-by-pixel after rasterisation, the boundary information is prepared in advance as part of the edge processing. This preliminary action eliminates the need for time-consuming post-rasterisation boundary extraction while maintaining accurate boundary definition.
3Loss of information
If the entire edge-based representation is rasterised to extract boundaries, then complete boundary information is obtained, but computational resources are wasted
Solution Approach 1:
The patent extracts only the necessary boundary information by introducing virtual edges at critical split points rather than rasterising the entire representation. The virtual edges are strategically placed at intersections and terminations of real edges, extracting only the boundary-defining elements needed for subsequent processing. This selective extraction obtains complete boundary information without the computational waste of full rasterisation.
Data Source
AI summary
A method receives an intermediate graphical representation (IGR) of a document, the IGR including a plurality of non-intersecting edges and determines a plurality of vertices using the IGR. The vertices include endpoints of the edges and a split point, being determined by splitting a first edge into two or more portions using an endpoint of at least one second edge, wherein the first edge and the second edge are from the plurality of non-intersecting edges, associates vertices from the plurality of vertices with at least one edge from the IGR using endpoints of the one edge, wherein the associating includes selecting, from the determined plurality of vertices, a plurality of neighbouring vertices to said one edge to define a region with the endpoints of the one edge. The method processes the region using at least the endpoints of said one edge and the associated neighbouring vertices.