Uses known inoculation positions and seeded watershed segmentation to separate peripheral target colonies from the central partner. The central fungus includes a dense core and connected peripheral hyphal growth. These masks are image-based tissue candidates, not inhibition boundaries. Morphological clean-up removes small fragments; ambiguous contact between organisms requires review. Target outlines are not forced into circles.
Usage
segment_colonies(
plate,
layout = plate_layout(),
min_area_mm2 = 0.5,
min_core_mm = 0.4,
morph_mm = 0.08,
hyphae_link_mm = 1,
edge_criterion = 0.5,
bridge_mm = 0.5
)Arguments
- plate
A plate processed by
classify_pixels().- layout
The
plate_layout()used for classification.- min_area_mm2
Minimum area of a fungal fragment (mm^2).
- min_core_mm
Minimum distance from core to colony edge (mm).
- morph_mm
Radius of morphological clean-up (mm).
- hyphae_link_mm
Maximum connected gap between central core and peripheral growth (mm).
- edge_criterion
Sub-pixel-consistent edge placement: within a 3-pixel band inside each colony outline, pixels are kept only if their colour difference from agar is at least this fraction of the colony's interior median ("half-maximum" criterion, unbiased for a blurred step edge). Set 0 to disable.
- bridge_mm
Opening radius (mm) separating central tissue from narrow connections.
Value
The plate with element objects: a list containing the colony
label matrix (labels, values = row index of colony_table),
satellites, bacteria, central_fungus, central_object and hyphae
masks, partner_type and colony_table. hyphae is connected peripheral central growth
candidate; its biological identity and boundary need QC.
Examples
sim <- simulate_plate(width = 300, height = 400, seed = 6)
p <- read_plate(sim$image) |> detect_plate() |> model_background() |>
classify_pixels(plate_layout()) |> segment_colonies(plate_layout())
p$objects$colony_table
#> id expected_x expected_y core_x core_y core_depth_mm detected
#> 1 TL 90.66136 140.6882 91 139 6.803420 TRUE
#> 2 TR 210.31293 140.6882 211 140 6.510846 TRUE
#> 3 BR 210.31293 260.3398 211 260 6.537189 TRUE
#> 4 BL 90.66136 260.3398 90 260 6.769656 TRUE
#> segmentation_trim_fraction
#> 1 0
#> 2 0
#> 3 0
#> 4 0