Example-based stippling using a scale-dependent grayscale process

Abstract

We present an example-based approach to synthesizing stipple illustrations for static 2D images that produces scale-dependent results appropriate for an intended spatial output size and resolution. We show how treating stippling as a grayscale process allows us to both produce on-screen output and to achieve stipple merging at medium tonal ranges. At the same time we can also produce images with high spatial and low color resolution for print reproduction. In addition, we discuss how to incorporate high-level illustration considerations into the stippling process based on discussions with and observations of a stipple artist. The implementation of the technique is based on a fast method for distributing dots using halftoning and can be used to create stipple images interactively.

Discrete element textures

“…Texture element placement Even though the majority of example-based texturing methods are not suitable for discrete elements, potential solutions have been explored by a few pioneering works, including 1D strokes [Jodoin et al 2002], 2D stipples [Kim et al 2009;Martín et al 2010], 2D particles/elements [Dischler et al 2002;Ijiri et al 2008;Hurtut et al 2009], and 2D agent motions [Lerner et al 2007;Ju et al 2010]. However, these methods treat each element as a single sample without a comprehensive neighborhood metric or a general synthesis solver as in our method; thus they might not faithfully reproduce both overall element distributions and individual element properties, especially shape, orientation, or heterogeneous elements.…”

Reducing affective responses to surgical images through color manipulation and stylization

“…ApparentGrey [Smith et al 2008] HueShift Bilateral [Tomasi and Manduchi 1998] Kuwahara [Kyprianidis 2011] ShapeSimpl [Kang and Lee 2008] CoherenceEnh [Kyprianidis and Kang 2011] XDoG [Winnemöller et al 2012] Oilpaint [Semmo et al 2016] Watercolor [Bousseau et al 2006;Wang et al 2014] BrushStrokes [Hertzmann 1998] Hatching [Praun et al 2001] Stippling [Martín et al 2010;Martín et al 2011] [Smith et al 2008] N = 4, p i = 0.5, k i = 0.5 HueShift custom, see Section 3.5 α = 120.0 Bilateral [Tomasi and Manduchi 1998] σ d = 3.0, σ r = 4.25% FlowAbs [Kyprianidis and Döllner 2008] ρ = 2, n e = 1, n a = 3, σ d = 6, σ r = 5.25%, σ e = 1, τ = 0.99, ϵ = 0, φ e = 2, σ m = 3, q = 8, φ q = 2 Kuwahara [Kyprianidis 2011] N = 8, ρ = 2.0, r = 6, q = 8, α = 1, τ w = 0.02, p s = 0.5, p d = 1.25, τ v = 0.1 ShapeSimpl [Kang and Lee 2008] N = 8, k = 8, r = 1, etf halfw = 3, etf N = 4, σ shock = 1, τ shock = 1 CoherenceEnh [Kyprianidis and Kang 2011] N = 4, σ d = 1, τ r = 0.002, σ t = 6, σ i = 0, σ д = 1.5, r = 2, τ s = 0.005, σ a = 1.5 XDoG [Winnemöller et al 2012] σ c = 2.28, σ e = 1, σ m = 4.4, p = 99.0, φ = 100.0, ϵ = 0.65, σ a = 1.0 Stylization Oilpaint [Semmo et al 2016] σ s = 12.0, n e = 0, σ b = 12.0, k scale = 5.0, k specular = 0.8, k shininess = 12.0 Watercolor [Bousseau et al 2006;Wang et al 2014] Effects of [Bousseau et al 2006;Wang et al 2014] using FlowAbs: ρ = 2.0, σ d = 4.0, σ r = 15.00% BrushStrokes [Hertzmann 1998] [Praun et al 2001] Art map of [Praun et al 2001] scaled at 0.6, linear mapping with luminance, edge settings of FlowAbs Stippling [Martín et al 2011] r es 0 = 1200 ppi, f p = 2.0, placement randomness = 25%, distribution = normal, colors = b&w, τ = 127…”

Weighted linde-buzo-gray stippling

“…Kim et al [2009] employ texture synthesis to create stipple textures from artists, which produces images with a more natural look. Martín et al [2010] use a halftoning method for creating initial stipple sets and later merge nearby stipples to reach a more uniform distribution. They focus on the stipples' shape and texture to produce an artistically looking output.…”