Detecting Viewer-Perceived Intended Vector Sketch Connectivity

(a) Free-hand vector line drawings are often imprecise with strokes intended to intersect stopping short of doing so; loops formed by raw strokes visualized on top left, each closed loop interior colorized with a different color, with the background left white. We successfully extract viewer perceived intended stroke connectivity distinguishing between intended junctions (a, e.g. circled in blue) and intended gaps (a, e.g. circled in red) (e) outperforming prior art (b, c). We arrive at this solution by combining local feature based predictions of the likelihood of pairs of strokes to form intended junctions (d) with global perceptual cues (e). Input image © The "Hero" artist Team under CC BY 4.0.

Abstract

Many sketch processing applications target precise vector drawings with accurately specified stroke intersections, yet free-form artist drawn sketches are typically inexact: strokes that are intended to intersect often stop short of doing so. While human observers easily perceive the artist intended stroke connectivity, manually, or even semi-manually, correcting drawings to generate correctly connected outputs is tedious and highly time consuming. We propose a novel, robust algorithm that extracts viewer-perceived stroke connectivity from inexact free-form vector drawings by leveraging observations about local and global factors that impact human perception of inter-stroke connectivity. We employ the identified local cues to train classifiers that assess the likelihood that pairs of strokes are perceived as forming end-to-end or T- junctions based on local context. We then use these classifiers within an incremental framework that combines classifier provided likelihoods with a more global, contextual and closure-based, analysis. We demonstrate our method on over 95 diversely sourced inputs, and validate it via a series of perceptual studies; participants prefer our outputs over the closest alternative by a factor of 9 to 1.

Paper

PDF (23.9 MB)

Downloads

Additional Results

BibTeX

@article{sketchconnectivity,
      title = "Detecting Viewer-Perceived Intended Vector Sketch Connectivity",
      author = {Yin, Jerry and Liu, Chenxi and Lin, Rebecca and Vining, Nicholas and Rhodin, Helge and Sheffer, Alla},
      year = 2022,
      journal = {ACM Transactions on Graphics},
      publisher = {ACM},
      address = {New York, NY, USA},
      volume = 41,
      issue = 4
}