scitex-cv

Small cv2-based image utilities — I/O, transform, filters, drawing.

Full Documentation · uv pip install scitex-cv[all]

Installation

pip install scitex-cv

Architecture

scitex_cv/
├── __init__.py        ← public API (load, save, resize, blur, edge_detect, ...)
├── _io.py             ← image I/O wrappers (PIL / OpenCV backends)
├── _transform.py      ← geometric transforms (resize, crop, rotate)
├── _filters.py        ← blur, sharpen, edge-detect filters
├── _draw.py           ← annotation helpers (draw boxes, text, masks)
└── _ocr.py            ← image → text (EasyOCR; optional `ocr` extra)

Thin, opinionated wrapper around PIL + OpenCV — every public name in __init__.py re-exports from one of the five leaf modules above.

Quick Start

import scitex_cv as cv

img = cv.load("input.png")
img = cv.resize(img, scale=0.5)
img = cv.blur(img, ksize=5)
edges = cv.edge_detect(img, method="canny")
cv.save(edges, "edges.png")

1 Interfaces

import scitex_cv as cv

# I/O
img = cv.load("input.png")
cv.save(img, "out.png")
gray = cv.to_gray(img); rgb = cv.to_rgb(img); bgr = cv.to_bgr(img)

# Transform
cv.resize(img, scale=0.5)
cv.rotate(img, angle=90)
cv.flip(img, direction="horizontal")
cv.crop(img, x=10, y=10, width=100, height=100)
cv.pad(img, top=10, bottom=10, left=10, right=10)

# Filters
cv.blur(img, ksize=5); cv.sharpen(img)
cv.edge_detect(img, method="canny")
cv.threshold(img, thresh=128); cv.denoise(img)

# Drawing
cv.rectangle(img, (5, 5), (30, 30), color=(0, 255, 0))
cv.circle(img, (50, 50), radius=10)
cv.line(img, (0, 0), (100, 100))
cv.arrow(img, (0, 0), (50, 50))
cv.text(img, "label", (10, 30))
cv.polylines(img, points=np.array([[0, 0], [100, 0], [100, 100]]))

# OCR (image → text; requires the optional `ocr` extra: pip install 'scitex-cv[ocr]')
cv.ocr("scan.png")                       # → recognized text (str)
cv.ocr(img, languages=("en",))           # in-memory array, English only
cv.ocr(img, detail=True)                 # → [(bbox, text, confidence), ...]

Demo

flowchart LR
    F["input.png"] --> L["cv.load()"]
    L --> R["cv.resize(scale=0.5)"]
    R --> B["cv.blur(ksize=5)"]
    B --> E["cv.edge_detect(method='canny')"]
    E --> S["cv.save('edges.png')"]
    S --> O["edges.png"]

Status

Standalone fork of scitex.cv. Only deps are numpy + opencv-python. The umbrella package's scitex.cv import path is preserved via a sys.modules-alias bridge.

Part of SciTeX

scitex-cv is part of SciTeX. Install via the umbrella with pip install scitex[cv] to use as scitex.cv (Python) or scitex cv ... (CLI).

Four Freedoms for Research

0. The freedom to run your research anywhere — your machine, your terms.
1. The freedom to study how every step works — from raw data to final manuscript.
2. The freedom to redistribute your workflows, not just your papers.
3. The freedom to modify any module and share improvements with the community.

AGPL-3.0 — because we believe research infrastructure deserves the same freedoms as the software it runs on.

License

AGPL-3.0-only (see LICENSE).