denissar / imageprocessing_course_icmc Goto Github PK

install virtualenvwrapper to manage virtual environments

pip install --user virtualenvwrapper

start the virtualenvwrapper

$ source /usr/local/bin/virtualenvwrapper.sh

make sure to include this at the shell session startup - usually .bashrc

echo 'source /usr/local/bin/virtualenvwrapper.sh' >> ~/.bashrc

in order to create a new environment called venv use:

mkvirtualenv venv

inside the virtual environment, install all needed python packages

(venv) $ pip install numpy
(venv) $ pip install imageio
(venv) $ pip install matplotlib
(venv) $ pip install scipy
(venv) $ pip install jupyter

if you want to enter the virtual environment later, just type

workon venv

it opens on your browser in order to create new notebooks or run existing ones

(venv) $ jupyter notebook

Folder Images contains images used in codes

1. Fundamentals: basic libraries, loading images, pixel processing, showing and writing images in disk

2. Enhancement

• (1) pointwise gray-level: image enhancement using pixel-wise gray-level values
• (2) histogram-based: using histograms to process images

3. Spatial Filtering (Convolution)

• Convolution: convolution operator, designing filters for spatial processing

4. Fourier Transform and Frequency Analysis

• (1) Frequency Analysis: introduction to the analysis of frequencies
• (2) Discrete Fourier Transform: the Discrete Fourier Transform in 1D and 2D
• (3) Fast Fourier Transform: the divide and conquer algorithm to implement Discrete Fourier Transform in O(n log n) time

5. Image Restoration

• (1) Denoising: simulating and studying noise, and filtering it out using denoising methods
• (2) Deconvolution: the point spread function and inverse problems

6. Colour images

• (1) Colour Image Processing: colour spaces and processing images
• (2) Colour Image Description: using colour information to extract features from images

7. Image Segmentation

• Threshold and Region-based Segmentation: strategies for threshold-based (histogram) and region-based (pixel) segmentation