Joint Angle Detection through Human Pose Estimation: A Proposal

Accurately detecting joint angles is fundamental in fields such as biomechanics, physical therapy, sports science, and human-computer interaction. This proposal introduces a framework for joint angle detection leveraging human pose estimation techniques. By utilizing deep learning models designed for pose estimation, we aim to identify key body landmarks and compute joint angles. The proposed method involves three key phases: (i) capturing human pose data using pre-trained pose estimation models, (ii) deriving joint coordinates, and (iii) computing angular relationships between the identified joints to sent notification related to incorrect postures. The proposed method provides a reliable, scalable, and non-invasive method for joint angle detection, paving the way for advancements in motion analysis, rehabilitation monitoring, and athletic performance optimization. The proposal holds potential for integration into wearable devices and real-time analysis systems, offering significant utility in both clinical and real-world scenarios, including augmented and virtual reality.