Please feel free to connect with me here on LinkedIn if you are interested in data science and machine learning.

Image processing examples with Numpy, Scipy, and Scikit-image

Requirements

• Python 3.4+
• NumPy ($ pip install numpy)
• SciPy ($ pip install scipy)
• MatplotLib ($ pip install matplotlib)
• Scikit-image ($ pip install scikit-image)

Testing after install

Open a Jupyter notebook and execute the following code,

import numpy as np
import matplotlib.pyplot as plt
from skimage import data, io, filters

image = data.coins()  # or any NumPy array!
edges = filters.sobel(image)
io.imshow(edges)

You should see the following output. If you see this, you are all set to go!

Simple NumPy array based operations

• Demo of NumPy based image manipulation
  

• Block view function and pooling/sampling
  

• Zooming (interpolation) based on SciPy

Exposure and color channel manipulations

• RGB to gray conversion
  

• RGB to HSV (Hue-Saturation-Value) conversion
  

• Adapting gray-scale filters to RGB images using special decorator

• Adjusting contrast by filtering regional maxima
  

• Local Histogram equalization
  

• Gamma and log contrast

• Tinting grayscale images
  

Edges, lines, and contours

• Finding contours
  

• Convex Hull of an image
  

• Skeletonize 2-D and 3-D images
  

• Marching cubes
  

• Edge operators
  

Geometrical transformations and registration

• Swirl
  

• Interpolation - edge modes
  

• Rescale, resize, downscale

• Histogram matching
  

• Structural similarity index

Filtering and restoration

Hysteresis thresholding

Image deconvolution

Unsharp mask