How can we improve images, estimate image gradients and combine images with filtering at different scales?
Lecture 3: Image processing
This week introduces image filtering and how we can process images at different scales.
Topics
Alternative video lectures in Norwegian
The videos above are new recordings performed in English on updated material. But last year's videos cover more or less the same material in Norwegian, for those who prefer that:
- Image filtering (alt. video)
- Image pyramids (alt. video)
- Laplace blending (alt. video)
Recommended reading
- Szeliski chapters 3.1-3.5.
Lab 3: Image blending with OpenCV
We are today going to play with image blending using Laplacian pyramids as explained in the lectures.
C++
Compared to the previous lab, this project is slightly more complex. The code is now distributed over a few more files to keep the logic parts of the program separated. If you are a beginner to programming and feel overwhelmed by this, take a look at section 1.2 in Introduction to C++ and OpenCV, which describes the concept of header files and source files. Section 2.2 covers the basics of loading images with OpenCV, which will also be useful in this lab.
You will find the lab at https://github.com/tek5030/lab-image-blending.
Follow the lab guide in README.md.
Python
You will find the lab at github.com/tek5030/lab-image-blending-py.
Follow the lab guide in README.md.