scikit-image — Scientific Image Processing

scikit-image — Scientific Image Processing

Overview

scikit-image is a Python library for image processing in the SciPy ecosystem. It provides algorithms for reading/writing images, filtering (noise reduction, edge detection), geometric transforms, segmentation (thresholding, watershed, active contours), object measurement (area, intensity, shape descriptors), and feature detection. Images are represented as NumPy arrays, enabling seamless integration with NumPy, SciPy, matplotlib, and pandas. Widely used for fluorescence microscopy, histology, and general bioimage analysis.

When to Use

• Preprocessing fluorescence microscopy images: background subtraction, denoising, illumination correction
• Segmenting cells, nuclei, or organelles using thresholding or watershed
• Measuring object properties: area, perimeter, intensity statistics, shape descriptors
• Applying morphological operations: erosion, dilation, opening, closing, fill holes
• Detecting keypoints or local features in biological images
• Converting between image formats and color spaces
• Use OpenCV instead for real-time video processing or GPU-accelerated operations
• For deep-learning cell segmentation, use CellPose instead (better accuracy for touching cells)
• Use napari instead for interactive multi-dimensional image visualization and annotation
• For whole-slide image tiling, use PathML or histolab instead

Prerequisites

• Python packages: scikit-image, numpy, scipy, matplotlib
• Input requirements: Images as files (TIFF, PNG, JPEG) or NumPy arrays; fluorescence images as 2D/3D grayscale arrays
• Environment: Python 3.9+

pip install scikit-image numpy scipy matplotlib

# For reading proprietary microscopy formats
pip install tifffile aicsimageio

# Verify
python -c "import skimage; print(skimage.__version__)"

Quick Start

from skimage import io, filters, measure
import numpy as np