countColors: Count the number of pixels within a color range or ranges
In countcolors: Locates and Counts Pixels Within Color Range(s) in Images

Description Usage Arguments Details Value See Also Examples

Counts the pixels within a color range; ranges can be spherical (see sphericalRange) or rectangular (rectangularRange). If multiple ranges are specified, each one is colored using a different indicator color.

countColors(path, color.range = "spherical", center, radius, lower,
  upper, bg.lower = rep(0.8, 3), bg.upper = rep(1, 3),
  target.color = c("magenta", "cyan", "yellow"), plotting = FALSE,
  save.indicator = FALSE, dpi = 72, return.indicator = FALSE)

`path`	Path to the image (JPEG or PNG).
`color.range`	Type of range being specified. Must be either "spherical" or "rectangular".
`center`	A vector or n x 3 matrix of color centers (RGB triplets) around which to search using spherical color range. RGB range 0-1 (not 0-255). See details.
`radius`	Values between 0 and 1 specifying the size of the area around `center` to search. The same number of centers and radii must be specified.
`lower, upper`	RGB triplet(s) specifying the bounds of color space to search. Must be the same length. See details.
`bg.lower, bg.upper`	RGB triplets specifying the bounds of color space to ignore as background, or `NULL` to use the entire image.
`target.color`	If an indicator image is created, the color with which to replace specified pixels. Can be either an RGB triplet or one of the colors listed by `colors`.
`plotting`	Logical. Should output be plotted in the plot window?
`save.indicator`	Logical OR path for saving indicator image. If TRUE, saves image to the same directory as the original image as 'originalimagename_masked.png'; if a path is provided, saves it to that directory/name instead.
`dpi`	Resolution (dots per image) for saving indicator image.
`return.indicator`	Logical. Should an indicator image (RGB array with targeted pixels changed to indicator color) be returned?

More than one set of ranges can be specified for the color search space, but the same number of arguments must be provided in each case (so the number of pixels and centers must be the same if using a spherical range, and the number of upper and lower bounds must be the same if using a rectangular one).

For center, upper, and lower, which call for RGB triplets, provide either a vector of RGB triplets in RGB order, i.e. c(R1, G1, B1, R2, G2, B2, etc) or a 3-column matrix with one row per RGB triplet. If a vector is provided, it is coerced to a 3-column matrix, and must therefore be a multiple of 3.

A list with the following attributes:

pixel.idx: Unique coordinates of pixels within color range(s).
pixel.fraction: Proportion of the non-background image within color range(s), found by dividing the number of pixels in pixel.idx by the number of non-background pixels in the image.
indicator.img: If return.indicator = TRUE, RGB array with color-swapped pixels.

countColorsInDirectory

# Try out a few different radii for white pelicans
pelicans <- system.file("extdata", "pelicans.jpg", package = "countcolors")

white.ctr <- rep(0.9, 9)
white.radii <- c(0.5, 0.3, 0.1)

# Magenta = 50% threshold, cyan = 30% threshold, yellow = 10% threshold
pelican.example <- countcolors::countColors(pelicans, center = white.ctr,
radius = white.radii, bg.lower = NULL, plotting = TRUE)