roi_rectangle

A small, typed helper for rectangular regions of interest (ROI) in NumPy images.

Install

pip install roi-rectangle

Quickstart

import numpy as np
from roi_rectangle import RoiRectangle

image = np.arange(100 * 120).reshape(100, 120)

roi = RoiRectangle(y1=30, y2=80, x1=20, x2=70)  # rows [30, 80), cols [20, 70)
print(roi.shape, roi.center, roi.area)          # (50, 50) (45, 55) 2500

crop = roi.slice(image)                         # same as image[30:80, 20:70]
print(crop.shape)                               # (50, 50)

roi2 = roi.recenter((60, 40)).resize(10, 20)    # methods return new ROIs
print(roi2)                                     # RoiRectangle(y1=30, y2=50, x1=55, x2=65)
print(image[roi2.get_slices()].shape)           # (20, 10)

API overview

RoiRectangle(y1, y2, x1, x2) is a dataclass. Intervals are half-open: [y1, y2) and [x1, x2). y2 or x2 may be None, meaning the region extends to the image edge on that axis. A ValueError is raised if y2 < y1 or x2 < x1.

Coordinates are passed to NumPy unchanged: an ROI past the bottom or right edge is clipped by NumPy, while negative coordinates count from the end of the axis.

Citation

If you use this package in your work, please cite it using the metadata in CITATION.cff.

License

MIT. See LICENSE.