fillLineGaps {biopixR}

Reconnecting discontinuous lines

Description

The function attempts to fill in edge discontinuities in order to enable normal labeling and edge detection.

Usage

fillLineGaps(
  droplet.img,
  bead.img = NULL,
  threshold = "13%",
  alpha = 0.75,
  sigma = 0.1,
  radius = 5,
  iterations = 2,
  visualize = TRUE
)

Arguments

droplet.img

image that contains discontinuous lines like edges or contours

bead.img

image that contains objects that should be removed before before applying the fill algorithm

threshold

"in %" (from threshold)

alpha

threshold adjustment factor for edge detection (from cannyEdges)

sigma

smoothing (from cannyEdges)

radius

maximal radius that should be scanned for another cluster

iterations

how many times the algorithm should find line ends and reconnect them to their closest neighbor

visualize

if TRUE (default) a plot is displayed highlighting the added pixels in the original image

Details

The function pre-processes the image to enable the application of adaptiveInterpolation. Pre-processing involves thresholding, optional object removal, LineEnd and diagonal LineEnd detection, and labeling. The threshold should be set to allow for some remaining "bridge" pixels between gaps to facilitate reconnection. For more details about reconnection, please consult adaptiveInterpolation. Post-processing involves thinning the lines. When removing objects from an image, their coordinates are collected using the objectDetection function. Next, the pixels of the detected objects are nullified in the original image, allowing the algorithm to proceed without the objects.

Value

image with continuous edges (closed gaps)

Examples

fillLineGaps(droplets)