Designing a Marker Based Motion Capture Setup for Sign Language Research

Motion capture systems designed for sign language must meet unique demands: capturing highly detailed hand, face, and body movements while preserving the linguistic and expressive integrity of the performance. This paper documents the iterative development of a marker-based motion capture setup tailored specifically for sign language data collection. Through a combination of hardware configuration, camera placement strategies, and workflow optimizations, we demonstrate how to balance tracking precision with signer comfort and recording efficiency. Our system enables the high-Throughput capture of individual signs-up to 500 per day-without relying on post-processing, thereby preserving the authenticity of the performance.Beyond presenting the workflow, this paper focuses on sharing insights gained throughout the development process, including the technical and practical challenges of working with a marker-based system. We discuss how issues such as marker size, facial tracking comfort, and infrared interference were addressed through iterative setup design. Currently, the system is optimized for recording isolated signs from a single, front-facing signer, which simplifies setup but limits the range of linguistic data that can be collected. Future work will focus on supporting multi-signer interactions, introducing new challenges in occlusion, camera coverage, and tracking fidelity.