Image Segmentation Program

Small Python program performing image segmentation using k-means clustering algorithm

Supported Image Types

• JPEG(.jpg)
• PNG(.png)

Python Libraries

• Scikit-Learn & Scikit-Image
  • Basic usage of k-means algorithm and image image loading.
• Numpy
  • Matrix operations
• Pillow(PIL)
  • Image convert
• Tkinter
  • GUI

Run the Program

• This can be done in Windows PowerShell or Linux/Mac Terminal, and it requires git, python, Numpy, PIL, Tkinter and Scikit, or at least pip.
• The instruction of how to install python can be found at HERE.
• The instruction of how to install pip can be found at HERE.
• The instruction of how to install git can be found at HERE.
• The Tkinter is included in python standard libraries and does not need specific installation once python is installed.
• To install numpy(it requires pip to be installed), run:

pip install numpy

• To install PIL(it requires pip to be installed), run:

pip install Pillow

• To install Scikit(it requires pip to be installed), run:

pip install -U scikit-learn

and

pip install scikit-image

• To start the program, sequentially run:

git clone https://github.com/RexWangSida/ImageSegmentationProgram.git

cd ImageSegmentationProgram

python frame.py

GUI

Examples of the image processing

Original Image

One Segment(One Cluster k == 1)

Two Segments(Two Clusters k == 2)

Three Segments(Three Clusters k == 3)

Four Segments(Four Clusters k == 4)

Five Segments(Five Clusters k == 5)

Six Segments(Six Clusters k == 6)

Seven Segments(Seven Clusters k == 7)

Eight Segments(Eight Clusters k == 8)

Nine Segments(Nine Clusters k == 9)

Ten Segments(Ten Clusters k == 10)