slambook is the companion code repository for the first edition of 14 Lectures on Visual SLAM: From Theory to Practice. It organizes practical C++ examples by book chapter so readers can compile, run, and modify each concept. The material progresses from Linux, CMake, rigid-body motion, Eigen, Lie groups, and camera models to complete SLAM components. It demonstrates nonlinear optimization with Ceres and g2o, feature-based visual odometry, and direct visual odometry. Later sections cover backend optimization, pose and factor graphs, loop closure, and dense reconstruction. A larger project directory combines concepts that are introduced separately in earlier chapters. The repository is educational rather than a ready-made production system and is distributed under the MIT License.
Features
- Chapter-organized C++ examples
- Visual SLAM mathematical foundations
- Eigen, Sophus, and OpenCV integration
- Ceres and g2o optimization exercises
- Feature-based and direct visual odometry
- Loop closure, graph optimization, and dense mapping