from __future__ import annotations
from typing import List, Literal
from phenotypic.abc_ import PrefabPipeline
from phenotypic.enhance import GaussianBlur
from phenotypic.detect import RoundPeaksDetector
from phenotypic.measure import (
MeasureShape,
MeasureIntensity,
MeasureTexture,
MeasureColor,
)
[docs]
class RoundPeaksPipeline(PrefabPipeline):
"""Detect and measure round colonies using lightweight peak-based detection.
A fast, minimal pipeline that applies Gaussian smoothing and grid-aware
peak detection for circular colonies. Fewer stages and parameters than
the Heavy variants, making it the fastest prefab option.
Steps:
1. GaussianBlur — smooth noise
2. RoundPeaksDetector — grid-aware circular colony detection
Measurements: MeasureShape, MeasureIntensity, MeasureTexture, MeasureColor.
Args:
blur_sigma: Gaussian blur sigma. Typical range: 1--5. Default: 5.
blur_mode: Boundary handling (``'reflect'``, ``'constant'``,
``'nearest'``). Default: ``'reflect'``.
blur_cval: Fill value when ``blur_mode='constant'``. Default: 0.0.
blur_truncate: Kernel extent in standard deviations. Default: 4.0.
detector_thresh_method: Thresholding method (``'otsu'``, ``'mean'``,
``'local'``, ``'triangle'``, ``'minimum'``, ``'isodata'``).
Default: ``'otsu'``.
detector_subtract_background: Normalize background before
thresholding. Default: ``True``.
detector_remove_noise: Morphological opening to remove specks.
Default: ``True``.
detector_footprint_radius: Radius for morphological operations.
Default: 5.
detector_smoothing_sigma: Sigma for grid profile smoothing.
Default: 2.0.
detector_min_peak_distance: Minimum grid line spacing. ``None``
auto-estimates. Default: ``None``.
detector_peak_prominence: Minimum peak prominence. ``None``
auto-estimates. Default: ``None``.
detector_edge_refinement: Refine grid edges using local profiles.
Default: ``True``.
texture_scale: Scale(s) for Haralick texture features. Default: 5.
texture_quant_lvl: Quantization level (8, 16, 32, 64).
Default: 32.
texture_enhance: Enhance contrast before texture measurement.
Default: ``False``.
texture_warn: Warn on unreliable texture measurements.
Default: ``False``.
benchmark: Enable per-step timing. Default: ``False``.
verbose: Enable verbose logging. Default: ``False``.
Best For:
- Well-separated round colonies on grid plates.
- High-throughput screening where speed matters.
- Plates with consistent colony sizes and regular spacing.
- Quick prototyping before switching to a heavier pipeline.
Consider Also:
- :class:`HeavyRoundPeaksPipeline` when additional refinement stages
are needed for cleaner results.
- :class:`HeavyOtsuPipeline` for general-purpose detection with more
robust preprocessing.
- :class:`FilamentousFungiPipeline` for filamentous organisms.
See Also:
:doc:`/tutorials/notebooks/08_using_prefab_pipelines` for a visual
comparison of prefab pipelines.
:doc:`/explanation/prefab_pipelines_guide` for guidance on choosing
a prefab pipeline.
"""
def __init__(
self,
*,
blur_sigma: int = 5,
blur_mode: str = "reflect",
blur_cval: float = 0.0,
blur_truncate: float = 4.0,
detector_thresh_method: Literal[
"otsu",
"mean",
"local",
"triangle",
"minimum",
"isodata",
] = "otsu",
detector_subtract_background: bool = True,
detector_remove_noise: bool = True,
detector_footprint_radius: int = 5,
detector_smoothing_sigma: float = 2.0,
detector_min_peak_distance: int | None = None,
detector_peak_prominence: float | None = None,
detector_edge_refinement: bool = True,
texture_scale: int | List[int] = 5,
texture_quant_lvl: Literal[8, 16, 32, 64] = 32,
texture_enhance: bool = False,
texture_warn: bool = False,
benchmark: bool = False,
verbose: bool = False,
) -> None:
gaussian = GaussianBlur(
sigma=blur_sigma,
mode=blur_mode,
cval=blur_cval,
truncate=blur_truncate,
)
detector = RoundPeaksDetector(
thresh_method=detector_thresh_method,
subtract_background=detector_subtract_background,
remove_noise=detector_remove_noise,
footprint_width=detector_footprint_radius,
smoothing_sigma=detector_smoothing_sigma,
min_peak_distance=detector_min_peak_distance,
peak_prominence=detector_peak_prominence,
edge_refinement=detector_edge_refinement,
)
texture_meas = MeasureTexture(
scale=texture_scale,
quant_lvl=texture_quant_lvl,
enhance=texture_enhance,
warn=texture_warn,
)
ops = [gaussian, detector]
meas = [
MeasureShape(),
MeasureIntensity(),
texture_meas,
MeasureColor(),
]
super().__init__(ops=ops, meas=meas, benchmark=benchmark, verbose=verbose)
__all__ = ("RoundPeaksPipeline",)
