enhanceHP: Enhance upward looking hemispherical photographs.

In GastonMauroDiaz/caiman: Canopy Image Analysis

Description

This algorithm uses the color perceptual attributes to enhance the contrast between the sky and plants through fuzzy classification. Color has three different perceptual attributes: hue, lightness, and chroma. The algorithm was developed using the following premise: the color of the sky is different from the color of plants. It performs the next classification rules, here expressed in natural language: clear sky is blue and clouds decrease its chroma; if clouds are highly dense, then the sky is achromatic, and, in such cases, it can be light or dark; everything that does not match this description is not sky. These linguistic rules were translated to math language by means of fuzzy logic.

Usage

enhanceHP(x, ...)

enhanceHemiPhoto(x, ...)

## S4 method for signature 'matrix'
enhanceHP(
  x,
  thr = NULL,
  fuzziness = NULL,
  wR = 0.5,
  wB = 1.5,
  skyBlue = colorspace::sRGB(matrix(normalize(c(135, 206, 235), 0, 255), ncol = 3)),
  ...
)

## S4 method for signature 'CanopyPhoto'
enhanceHP(
  x,
  mask = NULL,
  wR = 0.5,
  wB = 1.5,
  sharpen = FALSE,
  skyBlue = colorspace::sRGB(matrix(normalize(c(135, 206, 235), 0, 255), ncol = 3)),
  ...
)

Arguments

x	CanopyPhoto.
...	Additional arguments as for writeRaster.
thr	numeric. By default, the algorithm will try to estimate it.
fuzziness	numeric. By default, the algorithm will try to estimate it.
wR	numeric. Weight of red layer. See details.
wB	numeric. Weight of blue layer. See details.
skyBlue	color. See details.
mask	BinImage.
sharpen	logical. The default is TRUE, see details.

Details

This is a pixelwise algorithm that evaluates if pixels are sky blue colored. High score means high membership to Sky Blue. When a pixel are achromatic, then it uses pixel brightness. The algorithm internally uses membership2color and fuzzyLightness. The argument skyBlue is the targetColor of the former function, which output is the argument m of the latter function. To evaluate the brightness of an achromatic pixel, the algorithm uses Relative Brightness (see references).

Argument mask can be used to affect the estimation of two arguments of fuzzyLightness. Affected arguments are thr and fuzziness. The function autoThr is used to estimate thr. To compute fuzziness, the algorithm takes the maximum and the minimum values of the Relative Brightness and calculate its mean.

If sharpen is set as TRUE, a sharpen filter is applied to the raster with the membership values. This kernel is used: matrix(c(rep(-1, 3), -1, 12, -1, rep(-1, 3)), ncol = 3).

Value

numeric or RasterLayer.

Methods (by class)

• matrix: The output is a numeric vector.

• CanopyPhoto: The output is a RasterLayer.

References

Diaz, G.M., Lencinas, J.D., 2015. Enhanced Gap Fraction Extraction From Hemispherical Photography. IEEE Geosci. Remote Sens. Lett. 12, 1784-1789.

See Also

membership2color, fuzzyLightness.

Examples

# for circular hemispherical photos
x <- loadPhoto()
x <- normalize(x, 0, 255)
xe <- enhanceHP(x)
plot(xe)

z <- makeZimage(ncol(x), lensPolyCoef(c(0.6427, 0.0346, -0.024491)))

m <- doMask(z, zlim = asAngle(c(30,60)), alim = asAngle(c(0, 45)))
# mask affects the result but does not crop it (see Details).
plot(enhanceHP(x, mask = m))

GastonMauroDiaz/caiman documentation built on Jan. 22, 2022, 4:43 a.m.