Skip to contents

Convert Curvature Values to Smooth Gradient Colors

Source: R/view_surface.R
curv_cols_smooth.Rd

Maps surface curvature values to a continuous grayscale gradient, producing a FreeSurfer-style sulcal shading that is visually smoother than the binary mapping of curv_cols.

Usage

curv_cols_smooth(
  vals,
  light = "#C8C8C8",
  dark = "#4D4D4D",
  quantiles = c(0.05, 0.95),
  sharpness = 6
)

Arguments

vals

A numeric vector of curvature values for each vertex.

light

Hex color for the lightest shade (gyral crowns). Default "#D4D4D4".

dark

Hex color for the darkest shade (sulcal fundi). Default "#3A3A3A".

quantiles

Length-2 numeric vector of lower and upper quantiles used to clamp extreme values before rescaling. Default c(0.05, 0.95).

sharpness

Non-negative number controlling how crisp the gyral/sulcal split is. Larger values push vertices toward the light/dark extremes (a more binary, FreeSurfer-like contrast); smaller values give a softer gradient. 0 reproduces a plain linear ramp. Default 6.

Value

A character vector of hex color codes the same length as vals.

Details

Values are clamped to the range defined by quantiles to prevent outliers from washing out the colour map. The clamped values are then centred on their median (the gyral/sulcal boundary) and passed through a logistic contrast curve governed by sharpness before being interpolated between dark (sulcal fundi) and light (gyral crowns).

A plain linear ramp leaves most vertices near mid-grey, so the fold pattern tends to disappear once surface lighting is applied. Pushing values toward the two extremes with a logistic curve keeps the sulci clearly dark and the gyri clearly light, which reads as anatomical folding under lighting much the way FreeSurfer / Connectome Workbench curvature overlays do.

See also

curv_cols, curvature, view_surface

Examples

set.seed(1)
curv <- rnorm(500, sd = 0.1)
cols <- curv_cols_smooth(curv)
head(cols)
#> [1] "#5B5B5B" "#A1A1A1" "#545454" "#C8C8C8" "#ADADAD" "#545454"

# Softer, more continuous gradient
cols_soft <- curv_cols_smooth(curv, sharpness = 2)