Simultaneous Opto-Acoustic 3D Pose Estimation of Multiple Objects with Near-Far Mitigation for Economical Indoor Localization

Having knowledge about the positions and orientations of tools and workers' hands is highly beneficial in industrial assembly scenarios. This can be achieved by opto-acoustic indoor localization systems based on ultrasound and infrared that measure unilateral distances between mobile transmitters attached to objects (e. g. tools) and room-fixed receivers followed by multilateration for a 3D pose estimation. The operation in a code-division multiple access (CDMA) environment allows the tracking of multiple objects simultaneously. However, when two objects are having clearly different distances to one receiver, their code-modulated ultrasound and infrared signals arrive with considerably different intensities at the receiver due to attenuation. This complicates the identification of weak signals coming from objects being far away in the superimposed signal received and is referred to as near-far problem. Unfortunately, without near-far problem mitigation strategies, a correct unilateral distance measurement as foundation for a 3D pose estimation of far objects is not possible. Therefore, this contribution presents mitigation approaches: a step by step interference cancellation method and a subspace interference cancellation method that have not yet been applied to US and IR based indoor localization systems for position and orientation determination. Results of unilateral distance measurements and 3D pose estimation experiments of two objects in a near-far scenario are presented highlighting the superiority of the subspace interference cancellation method.