Digital image processing in OpenCV using Python

This repository shows usage of OpenCV Python package for digital image processing.

Linux Setup

1. Clone the repository:

git clone https://github.com/ackermannW/image-processing-CV
cd image-processing-CV

2. Run the bash setup script setup.sh. This script downloads miniconda, installs it and creates a python environment from the environment.yml file. When prompted accept the T&C by pressing a.

chmod a+x ./setup.sh
./setup.sh

3. Activate the environment by using

conda activate image-processing-cv

Windows Setup

1. Make sure that GIT is installed.

2. Clone the repository:

git clone https://github.com/ackermannW/image-processing-CV
cd image-processing-CV

3. Run the powershell setup script setup.ps1. This script downloads miniconda, installs it and creates a python environment from the environment.yml file. When prompted accept the T&C by pressing a.

.\setup.ps1

Note that it is also recommended to utilize Windows subsystem for Linux on Windows OS.

Contents

1. Basic image transformations
2. Image space filtering
3. Image filtering in frequency domain
4. Edge detection
5. Histogram and thresholding
6. Image segmentation
7. Classification and object detection

GPU setup

It is recommended to utilize GPU acceleration to speed up deep learning tasks. Instructions for CUDA cores can be found here in the official Tensorflow documentation. Note that it is recommended to use WSL on Widnows OS.