Large Scale Indoor Camera Positioning

This project tackles the task of estimating the pose of a large set of fixed indoor cameras, essential for applications such as augmented privacy, autonomous navigation, video surveillance, and logistics. Our innovative approach uses fiducial markers to initially estimate the pairwise relationship between nearby cameras and then uses a full optimization process that incorporates real-world information to refine these estimates. The method is validated through tests on both artificially created and real-world datasets, showing improved performance over existing methods.

This project involves two main types of datasets available at: https://sourceforge.net/projects/indoor-camera-positioning-data/.

Developed using CPP , CMake and QT Creator . We include a user-friendly GUI Interface.