RootscanR (version 0.0.1)

createDepthLayerMasks: Create masking layers for the depth layers

Description

createDepthLayerMasks - Creates a masking layer, i.e., a true/false-matrix, indicating which pixels of an image belong to a specified depth layer.
We assume the image was taken inside a minirhizotron, a round plexiglas tube used to observe roots in the ground, and shows a 360 degree scan of a section of the tube. As the minirhizotron is installed at an angle (can be specified, by default 45 degrees) a depth layer in the ground is shown as a sinus curve in the scan.

dissectionLine - Returns the y-value of the dissection line of a tube that was positioned at an angle for a given position x. The dissection line is dependent on the minpoint, the angle, and the radius of the tube (have to be specified).

Usage

createDepthLayerMasks(
  ppcm = NULL,
  ppi = NULL,
  dims_px,
  depth_levels_cm = rbind(c(0, -10), c(-10, -20), c(-20, -30), c(-30, -40)),
  depth_highpoint_cm = 0,
  pos_highpoint_px = "center",
  angle = 45,
  top_side = "top",
  gap_cm = 0,
  gap_deg = 0
)
dissectionLine(x, minpoint, angle, radius)

Value

createDepthLayerMasks A list of 2-dimensional true/false-matrices with the specified dimensions (see dims). True indicates that the corresponding pixel belongs to the specified depth layer. The list contains nrow(depth_levels_cm)-many matrices, one for each specified depth layer.

dissectionLine

Arguments

ppcm

Numeric value specifying how many pixels there are per centimeter (resolution of the image). Default NULL. If ppcm is not NULL or NA, ppi is ignored.
Examples: 300 ppi (printing) is 118 ppcm, 150 ppi is 59 ppcm, 72 ppi (screens) is 28 ppcm.

ppi

Numeric value specifying how many pixels there are per inch (resolution of the image). Default NULL. Leave/set ppcm to NULL or NA to use ppi.
Examples: 300 ppi (printing) is 118 ppcm, 150 ppi is 59 ppcm, 72 ppi (screens) is 28 ppcm.

dims_px

Integer vector of length two specifying the dimensions of the image/mask layer (in pixels).

depth_levels_cm

Numeric matrix with two columns and at least one row. Each row specifies a depth interval (in cm) that is of interest. For each interval a masking layer is generated. The first value of each row has to be larger than the second (decending).
By default: Depth layers Depth layers 0-(-10)cm, -10-(-20)cm, -20-(-30)cm, and -30-(-40)cm.

depth_highpoint_cm

Numeric value specifying the depth at which the highest point of the image is (in cm). By default this is at depth 0.
More precisely, this refers to the depth of the middle of the pixel in the first row of the top_side of the image.

pos_highpoint_px

Either one of "center" or "edge" (default), indicating that scanning starts at the bottom or top facing side of the minirhizotron, respectively, or it can also be an integer value (>=1 and <=width of the top_side of the image) specifying where (left<->right) in the top row of the image the highest point of the image lies, indicating the column where the minimum points of the depth-level lines lie.
"edge" is equivalent to using 1 or width of the top_side and "center" to using half the width of the top_side.
See also gap_cm if there is a non-zero gap.

angle

Numeric value >=0 and <=90 (default 45) specifying the installation angle of the minirhizotron in degrees (angle between the ground and the tube above the soil).

top_side

One of "left","top" (default),"right","bottom". Indicates where the upper side of the image is.

gap_cm

Numeric value (default 0) that specifies if there is a gap (in cm) between the two "matching" sides of the image, i.e., there is no full rotation of the scanner. If pos_highpoint_px is specified as "center" or "edge" (in characters), then it is assumed that the middle of the gap is right at the top or bottom facing side of the tube, i.e. it is not possible to start and end at the exact top/bottom. If pos_highpoint_px is specified as a numeric value, then it is considered as exact and the gap does not move it.

gap_deg

Numeric value (default 0) that specifies if there is a gap (in degrees, i.e. from 0 to 360) between the two "matching" sides of the image, i.e., there is no full rotation of the scanner. Works similar to gap_cm. Example: If the scanner scans only 355 degrees, then set the value to 5. gap_deg is only used if gap_cm is NA.

x

Numeric value indicating the position.

minpoint

Numeric vector of length 2 indicating the coordinates of the minimum point of the function.

radius

Positive numeric value indicating the radius of the tube.

Examples

# Small example for creating a masking layer of a single depth interval:
createDepthLayerMasks(ppcm = 1, dims_px = c(10,10),
                      depth_levels_cm = matrix(c(-3,-5), ncol = 2))[[1]]
# Change some parameters to adapt to a different minimum point position and
# installation angle.
createDepthLayerMasks(ppcm = 1, dims_px = c(5,7),
                      depth_levels_cm = matrix(c(-1.5,-2.5), ncol = 2),
                      pos_highpoint_px = "center", gap_cm = 0)[[1]]
# And here is a cutout of that mask using a gap.
createDepthLayerMasks(ppcm = 1, dims_px = c(5,5),
                      depth_levels_cm = matrix(c(-1.5,-2.5), ncol = 2),
                      pos_highpoint_px = "center", gap_cm = 2)[[1]]
# Examples how different angles result in different dissection graphs:
test_x <- -12:12
plot(test_x, sapply(X=test_x, function(X){
               dissectionLine(x=X, minpoint = c(0,0), angle = 45,
                              radius = 3)}),
     ylim = c(0, 20))
plot(test_x, sapply(X=test_x, function(X){
               dissectionLine(x=X, minpoint = c(0,0), angle = 80,
                              radius = 3)}),
     ylim = c(0, 20))
plot(test_x, sapply(X=test_x, function(X){
               dissectionLine(x=X, minpoint = c(0,0), angle = 10,
                              radius = 3)}),
     ylim = c(0, 20))