Skip to contents

Convert image to binary mask using flood fill

Source: R/masks.R
img_mask.Rd

Creates a binary mask by flood-filling the background from top-left corner, then setting all non-background pixels to black (TRUE). This handles complex backgrounds and multiple objects uniformly.

Usage

img_mask(img, fuzz = 20, invert = TRUE)

Arguments

img

Integer matrix (0-255) from img_load()

fuzz

Integer. Fuzz/tolerance for flood fill (0-255). Default 20 Controls how similar pixels must be to be considered background.

  • Low (5-10): Very strict, only very similar pixels

  • Medium (20-30): Reasonable for most cases

  • High (50+): Aggressive, may include object edges

invert

Logical. Invert the mask? Default TRUE

  • TRUE: object = FALSE (white), background = TRUE (black)

  • FALSE: object = TRUE (black), background = FALSE (white)

Value

Logical matrix (TRUE/FALSE depending on invert setting)

Examples

if (FALSE) { # \dontrun{
# Standard (inverted): white object on black background
img <- img_load("leaf.png")
mask <- img_mask(img)

# Non-inverted: black object on white background
mask_std <- img_mask(img, invert = FALSE)

# Adjust fuzz for noisy backgrounds
mask <- img_mask(img, fuzz = 30)
} # }