State of the art identification and positioning of other users in augmented reality applications

The current AR implementations mainly offer visual augmentation of the scenes surrounding the users. Users view their environment through an interface, such as a smartphone or a headset, and visual digital content is overlaid on top of the real-world view, providing informational and interactive features to an experience. In multi-user experiences, visual information that is relevant to each individual is displayed, and the users are free to interact in the added digital layer. For this to be possible, the device must be able to identify and position other users in its field of view. In AR games, this enables multiplayer features, while in industrial AR, it provides situation-specific and task-specific assistance to workers, enhancing efficiency.

The most common approach is to provide identification and positioning through a combination of:

• Visual Positioning System (VPS), applying AI-powered algorithms to recognize the environment, based on the image stream from the camera. VPS is the workhorse that provides the accuracy needed to enable high quality AR experiences, including multi-user.
• Global navigation satellite system (GNSS), such as GPS, Galileo, BDS and GLONASS. This approach is mainly supportive, working better the larger the distance to the object is.
• Inertial sensors on the device, including accelerometers, gyrometers and magnetometers. These sensors enable the positioning of the observer and provides useful data to compensate for their movements.