Particle filter-based camera localisation using square fiducials for augmented reality applications
Résumé
This paper describes a new approach for estimating the camera pose using Particle Filter framework (PF) which is well adapted for augmented reality applications. The proposed method can robustly track the camera position and orientation using only point features defined from reference square fiducials. Given extracted image points and the previous state, the PF allows to update iteratively the camera 3D motion parameters. Results from real data in an augmented reality set-up are presented demonstrating the efficiency and robustness of the proposed method.