A fiducial marker system using digital techniques

A FIDUCIAL MARKER SYSTEM USING DIGITAL TECHNIQUES FULL

To avoid rotation ambiguities, TopoTag introduces baseline node in its topological structure. High robustness with topological design helps saving tag bits for encoding identities. Additional bits will also be reserved for incorporating Hamming distance strategy in order to improve false positive rejection. Existing fiducial marker systems, especially square patterns, sacrifice tag bits to handle four rotation ambiguities during decoding. TopoTag utilizes topological structure information in tag design, which has been proven with robustness with illuminate variation and false detection. As a contrast, TopoTag offers more feature correspondences for a better pose estimation. Square markers choose to use four corner points for pose estimation, which is the minimum number for unambiguous pose estimation. ChromaTag uses color over AprilTag to improve marker detection speed. propose ArUco using mixed integer programming to generate markers. AprilTag is a faster and more robust reimplementation of ARTag. Several works investigate using multiple fiducial markers in a checkerboard to improve camera calibration and reduce the perspective ambiguity by further adding color. propose using nested markers to improve performance under different distances. propose using SIFT and SURF filters that are specifically designed for SIFT and SURF detectors. BinARyID proposes a method to generate markers focused on avoiding rotation ambiguities. In addition, they are designed with error correction mechanism to increase robustness. ARTag and ARToolkitPlus improve the recognition technique with a binary coded pattern. It includes a pattern in their inner region for identification via image correlation. ARToolkit is well known and widely used in many virtual reality applications.

Matrix, CyberCode and VisualCode are the first and simplest proposals. To be easily localized, the majority of fiducial systems are designed to contain a thick square border. Our dataset, marker design and detection algorithm are public to the community 1 1 1TopoTag homepage. In addition, TopoTag supports occlusion as long as main tag topological structure is maintained and flexible shape design where users can customize inter and outer marker shapes. Experiments show that TopoTag significantly outperforms previous fiducial marker systems in terms of various metrics, including detection accuracy, vertex jitter, pose jitter and accuracy, etc.

We collect a large dataset including in total 169,713 images for evaluation, involving in-plane and out-of-plane rotation, image blur, different distances and various backgrounds, etc.

A FIDUCIAL MARKER SYSTEM USING DIGITAL TECHNIQUES FULL

Without sacrificing bits for higher recall and precision like previous systems, TopoTag can use full bits for ID encoding and supports tens of thousands unique IDs and easily extends to millions and more by adding more bits, thus achieves perfect scalability. TopoTag uses topological and geometrical information in marker detection to achieve higher robustness. We introduce TopoTag, a robust and scalable topological fiducial marker system, which supports reliable and accurate pose estimation from a single image. Fiducial markers have been playing an important role in augmented reality (AR), robot navigation, and general applications where the relative pose between a camera and an object is required.

A fiducial marker system using digital techniques