Abstract:
Precise monitoring of the human motion and its construction in 3D is a process employed in many fields. It can be used in monitoring the performance of a sport person – measuring the angles between their upper and forearm while they are swimming or throwing a javelin or when they’re bowling. It can be used to ease the process of physiotherapy by providing a physiotherapist with the necessary equipment to continuously monitor the muscular movement of their patient thereby reducing the risk of increasing the injury. It can be used in training a person to let’s say, shoot an arrow.
Many devices have been introduced that have the ability to monitor and reproduce the human motion in 3D. However, these devices are expensive, need a strictly controlled environment or time consuming and repetitive processes of calibration and are prone to errors due to self-occlusion and the fields present in the environment.
D3D is a novel 3D motion reconstruction system that can improve the monitoring and reproduction process. Miniaturized multiple cameras embedded in wearable sleeve or vest are connected to a microprocessor to record short videos of the human motion. This relative motion is processed through an algorithm called Structure-from-Motion (SFM) to yield the position of cameras which in turn is used to reconstruct the 3D model of the arm or body’s motion in MATLAB and MESHLAB. The reconstructed motion can then be compared to the ideal motion and error can be determined accordingly as per the device’s usage.
