Fast Artificial Landmark Detection for Indoor Mobile Robots

Nowadays the big challenge in simultaneous localization and mapping (SLAM) of mobile robots is the creation of efficient and robust algorithms. Significant Number of SLAM algorithms rely on unique features or or use artificial landmarks received from camera images. Feature points and landmarks extraction from images have two significant drawbacks: CPU consumption and weak robustness depending on environment conditions. In this paper we consider performance issues for landmark detection, introduce a new artificial landmark design and fast algorithm for detecting and tracking them in arbitrary images. Also we provide results of performance optimization for different hardware platforms.