Analyse one plate photograph — analyze

Runs the complete FunMel pipeline: import, dish detection and scale, optional colour calibration, flat-field correction, colour/texture pixel classification, seeded colony separation, measurement and QC.

Usage

analyze_plate(
  path,
  layout = plate_layout(),
  id = NULL,
  dish_diameter_mm = 90,
  agar_fraction = 0.88,
  dish = NULL,
  mm_per_px = NULL,
  reference = NULL,
  reference_lab = c(50, 0, 0),
  max_dim = 2000,
  flat_field = TRUE,
  max_clipped_fraction = 0.02,
  classifier = NULL,
  min_delta_e = 3,
  min_area_mm2 = 0.5,
  min_core_mm = 0.4,
  morph_mm = 0.08,
  edge_criterion = 0.5,
  edge_exclude_mm = 0.1,
  dark_L = NULL,
  n_boot = 200L,
  block_mm = 0.5,
  ring_width_mm = 0.5,
  profile_mode = c("absolute", "relative"),
  n_rings = 10L,
  qc_file = NULL,
  keep_images = TRUE,
  verbose = TRUE,
  n_spots = NULL,
  spot_radius_mm = 0.5
)

Arguments

path: Image file or H x W x 3 sRGB array.
layout: A plate_layout(). Use plate_layout(centre = "none") for control plates without a central partner.
id: Plate identifier (default: file name).
dish_diameter_mm: Outer diameter of the dish (mm); sets the scale.
agar_fraction: Fraction of the dish radius that is analysed.
dish: Optional manual dish circle c(x, y, r) in working pixels.
mm_per_px: Optional known scale (mm per working-image pixel); overrides dish_diameter_mm.
reference: Optional exposure / white-balance reference: a grey-card region c(x, y, r) in relative image coordinates, or "agar"; see model_background().
reference_lab: Known CIELAB of the reference.
max_dim: Working resolution (longest side, px).
flat_field: Request illumination correction (recommended). It is applied only if the fitted correction passes the max_clipped_fraction check; plate_summary$flat_field_status records the outcome.
max_clipped_fraction: Largest fraction of dish pixels whose fitted flat-field gain may saturate before the correction is withheld as degenerate; see model_background(). Inf restores unconditional correction.
classifier: Optional supervised classifier (train_classifier()).
min_delta_e: Minimum colour difference from agar for objects.
min_area_mm2, min_core_mm, morph_mm, edge_criterion: Segmentation settings, see segment_colonies().
edge_exclude_mm, dark_L, n_boot, block_mm: Measurement settings, see measure_colonies().
ring_width_mm, profile_mode, n_rings: Profile settings, see colony_profiles().
qc_file: Optional path of a PNG quality-control figure.
keep_images: Keep image arrays in the result (needed for plot_qc() later; set FALSE in large batches to save memory).
verbose: Print progress.
n_spots: Optional spots per target colony; NULL keeps only whole-colony measurements.
spot_radius_mm: Circular spot radius (mm); see sample_colony_spots().

Value

An object of class funmel_result, a list with

colonies: One row per expected colony (measure_colonies()).
spots: Optional circular ROI measurements and locations; whole-colony MI stays unchanged.
profiles: Edge-to-centre profiles (colony_profiles()).
plate_summary: One row of plate-level measures and QC.
plate: The processed plate (images and masks) if keep_images = TRUE.
params, warnings: Settings used and QC warnings.

Examples

sim <- simulate_plate(width = 300, height = 400, seed = 10)
res <- analyze_plate(sim$image, id = "sim10", verbose = FALSE)
res
#> <funmel_result> plate "sim10"
#> • Scale: 0.2393 mm/px; agar L* = 22.2
#> • Colonies detected: 4 / 4; satellites: 1
#> • Central fungal/partner extent: equivalent r = 17.11 mm
#>  colony_id area_mm2 L_mean MI_mean MI_lo MI_hi delta_MI_facing
#>         TL      155   51.2    48.8  48.4  49.1            2.30
#>         TR      155   47.0    53.0  52.7  53.3            2.49
#>         BR      153   40.0    60.0  59.5  60.4            2.55
#>         BL      156   57.7    42.3  42.0  42.6           -2.57
#>  growth_inhibition_pct
#>                 3.5672
#>                 0.0496
#>                -0.3681
#>                -7.7772
res$colonies[, c("colony_id", "MI_mean", "delta_MI_facing")]
#>   colony_id  MI_mean delta_MI_facing
#> 1        TL 48.75959        2.299796
#> 2        TR 53.03797        2.491872
#> 3        BR 59.95116        2.551580
#> 4        BL 42.25989       -2.565176
if (FALSE) { # \dontrun{
res <- analyze_plate("plates/2026_05_26_20.JPG", layout = plate_layout())
plot_qc(res)
ctrl <- analyze_plate("plates/2026_05_26_37.JPG",
                      layout = plate_layout(centre = "none"))
} # }