Answers are grounded in this paper and cite their sources.
Patch‐based image vectorization methods usually represent image regions using fixed‐sided patches, such as triangular or quadrilateral patches. However, such patches are often inefficient for irregular or semantically complex regions, as many small patches are required to approximate them accurately. To solve this problem, we propose an image vectorization method based on generalized Coons patches, representing an input image as a set of shape‐adaptive polygonal patches. Compared with methods based on conventional Coons patches, our method can represent images with fewer patches and substantially alleviate fold‐over problems in irregular regions. To further improve the reconstruction accuracy inside each patch, an error‐driven adaptive refinement mechanism is introduced. Interior control points are directly inserted at locations with large reconstruction errors, and their color attributes are determined through color fitting, without re‐partitioning patches or modifying the patch topology. This mechanism is simple and flexible, and can progressively improve image reconstruction quality according to a prescribed error threshold. Experimental results demonstrate that our method achieves high‐quality image reconstruction on various input images.
No takes yet. Share an insight, caveat, or question.
Jing et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: