cv_algorithms

A Python package (Python3 ready!) that contains implementations of various OpenCV algorithms are are not available in OpenCV or OpenCV-contrib. This package is intended to be used with OpenCV 3.

Some performance-critical algorithms are written in optimized C code. The C code is accessed using cffi Currently implemented:

• Morphological algorithms
  • Guo-Hall thinning (C-optimized)
  • Zhang-Suen thinning (C-optimized)
  • Difference-of-Gaussian transform
• Algorithms on contours
  • Masking extraction of convex polygon area from image without rotation
  • Scale around reference point or center
  • Fast computation of center by coordinate averaging
  • Center-invariant rescaling of upright bounding rectangle by x/ factors
  • Filter by min/max area
  • Sort by area
  • Create binary contour mask
  • Grassfire transform
• Colorspace metrics & utilities:
  • Convert image to any colorspace supported by OpenCV
  • Extract any channel from any colorspace directly
  • Euclidean RGB distance
• Other structural algorithms
  • Which neighboring pixels are set in a binary image?
• Algorithms on text rendering
  • Center text at coordinates
  • Auto-scale text to fix into box
• Other algorithms
  • Remove n percent of image borders
  • Popcount (number of one bits) for 8, 16, 32 and 64 bit numpy arrays
  • Resize an image, maintaining the aspect ratio

As OpenCV's Python bindings (cv2) represents images as numpy arrays, most algorithms generically work with numpy1 arrays.

Installation

Install the stable version:

pip install cv_algorithms

How to install the bleeding-edge version from GitHub

pip install git+https://github.com/ulikoehler/cv_algorithms.git

How to build yourself - we use Poetry:

poetry build

Potentially, you need to install OpenCV if not already present. I recommend first trying to install without that, since modern Python versions will take care of that automatically.

Usage

Difference of Gaussian transform documentation & example Grassfire transform documentation & example Thinning documentation & example

Here's a simple usage showcase:

import cv_algorithms
# img must be a binary, single-channel (grayscale) image.
thinned = cv_algorithms.guo_hall(img)

Contributions

Contributions of any shape or form are welcome. Please submit a pull request or file an issue on GitHub.

Copyright (c) 2016-2022 Uli Köhler <cv_algorithms@techoverflow.net>