Smithangshu / ImageAlignmentCpp

C++ Implementation of Baker-Matthews Inverse Compositional Image Alignment algorithm, with speed and accuracy optimisations

Baker-Matthews Image Alignment in C++, with speed and accuracy optimisations

This C++ application aims to implement the Baker-Matthews inverse compositional image alignment algorithm, with a robust M-estimator to add robustness against lighting conditions.

This C++ implementation is an adaption of Assignment 7 of Carnegie Mellon University's http://16-385 Computer Vision course in Fall 2020. It is a great course: I advise CMU students to take it if possible. Lecture slides from which this implementation is derived can be found publicly on the course website

Running the Application

0. Clone the repository

git clone https://github.com/Samleo8/ImageAlignmentCpp

1. Install dependencies

See below.

2. Build with cmake

cd build
cmake ..

3. Make and run the application

make
./TestKLT

Dependencies

Eigen (min 3.3.7)

OpenCV (suggested 4.5.*)