I've set up this repo to dig deeper into image processing algorithms, using C++ and the CImg Library. College courses only scratch the surface of what's out there, so this project is my way of diving in and getting hands-on with C++.
I'm using the book Digital Image Processing with C++: Implementing Reference Algorithms with the CImg Library by Tschumperlé, Tilmant, Barra as my primary reference.
Each script in this repository is designed to run independently to maximize modularity. While this may not adhere to traditional software design principles, the focus here is on isolated, hands-on learning of individual image processing algorithms.
I used to rely on OpenCV for image processing tasks. It's undoubtedly powerful, but getting it set up on a new computer was always a bit of a headache. Also, it has so many features that it was hard to keep track of everything.
Then I discovered CImg, a streamlined library that encapsulates everything in a single header file. That means you can just drop it into your project and get going. I'll be sharing what I learn as I go along, so check out my study notes if you're interested.
Here are some key resources that guided this project:
- Primary Reference: Digital Image Processing with C++: Implementing Reference Algorithms with the CImg Library by Tschumperlé, Tilmant, Barra
- CImg Library
- Principles of Digital Image Processing series by Burger & Burge (2009, 2013)
| Chapter | Topic | Key Ideas |
|---|---|---|
| 1 | Getting Started | Installation, hello world, and CImgDisplay |
2  |
Block Decomposition | CImg template class, reading command-line parameters, get vs. non-get methods, block decomposition, loops in CImg, draw methods |
3  |
Point Processing Transformations | Mathematical transformations, bitwise transformations, histogram equalization |
4  |
Mathematical Morphology | Dilation, erosion, opening, closing, Kramer-Bruckner, Alternating Sequential Filters (ASF), morphological gradients, Beucher gradient, skeletonization |
5  |
Filtering | Convolution, median filter, first-order derivative, second-order derivative, adaptive filters, adaptive window filters, frequency-domain filtering, diffusion filtering |
6  |
Feature Extraction | Harris-Stephens corner detection, Shi-Tomasi algorithm, Hough transform, texture spectrum, Tamura texture coefficients, Local Binary Patterns (LBP) |
7  |
Segmentation | Active contours, Ostu's algorithm, Bernsen's algorithm, K-means clustering, Simple Linear Iterative Clustering (SLIC) |
8  |
Motion Estimation | Horn-Schunck optical flow, multiscale analysis, Lucas-Kanade optical flow, eigenelement analysis, cross-correlation, phase correlation, Kalman filter |
9  |
Multispectral Image Processing | Principal Component Analysis (PCA), median filtering in color images, color spaces |
10  |
3D Visualization | Creating simple 3D shapes |
| Appendix 1 | Math Expressions in CImg's Fill Method | How to write math expression for CImg<>::fill() |
| Appendix 2 | JPEG Compression | JPEG algorithm |