Learn-scikit-image

This repo is to learn some image processing in python by scikit-image module.

setup

Windows for conda-based distributions

$ conda install scikit-image

Linux and OSX

$ pip install -U scikit-image

module

skimage.io

• io.imread(URL or xxx.jpg) : Load an image from file.
• io.imshow(img) : Display image
• io.imsave(NAME) : Save an image to file.
• io.ImageCollection : Load a collection of images.

skimage.color

Do the color space conversion such as: 1. RGB to HSV color space conversion. (rgb2hsv(img)) 2. RGB to XYZ color space conversion. (rgb2xyz(img)) 3. RGB to RGB CIE color space conversion. (rgb2rgbcie(img)) 4. CIE-LAB to CIE-LCH color space conversion. (lab2lch(img))

You can also use convert_colorspace(img, fromspace, tospace) to convert color space.

skimage.exposure

• Histogram equalization : Make the intensity distribution better of image. It can get higher contrast.

  • Histogram equalizatio ---wiki
  • exposure.histogram : histogram of image
  • exposure.equalize_hist : image after histogram equalization

• Gamma Correction : In human's vision, the brighteness to human is not in nonlinear. We should use gamma correction to make the brightness much linear in human sight.

  • Gamma Correction ---wiki
  • exposure.adjust_gamma(img, gamma_value) : Adjust gamma_value for image
  • exposure.is_low_contrast(img) : To justify whether the image is in low contrast

skimage.data_dir

• data_dir : It is sample data directory under the absolute path where you setup scikit-image