from typing import cast
import numpy as np
from PIL import Image
from skimage import filters
from .base import TissueDetector
[docs]
class OtsuTissueDetector(TissueDetector):
"""Otsu-based tissue detection
This detector applies Otsu thresholding with optional morphological operations.
It's fast, robust, and works well for standard histopathology images.
"""
[docs]
def detect(self, image: Image.Image) -> np.ndarray:
"""Detect tissue using Otsu thresholding.
Parameters
----------
image : Image.Image
Input RGB image
Returns
-------
np.ndarray
Binary mask (dtype: uint8 with values 0 or 1)
"""
# Convert to numpy and grayscale
img_array = np.array(image.convert("L")).astype(np.uint8)
# Apply Otsu thresholding
threshold = cast(float, filters.threshold_otsu(img_array)) # type: ignore
tissue_mask: np.ndarray = (img_array < threshold).astype(np.uint8)
return tissue_mask
GlassCut