A wearable motion capture device able to detect dynamic motion of human limbs

A wearable motion capture device able to detect dynamic motion of human limbs

Human motion capture using wearable devices is essential in action recognition, motor function assessment and dexterous human-robot interaction for rehabilitation robots and intelligent prosthetics. It allows machine to assist users and improve life quality in senior care, physical rehabilitation, daily life-logging, personal fitness, and assistance for people with cognitive disorders and motor dysfunctions. Unfortunately, existing wearable devices usually suffer from drift and instability problems in capturing highly-dynamic motion of limb activities. How to accurately and robustly detect limb motions using a low-cost and simple device is still a great challenge. In our recent work  (Nature Communications volume 11, 5615, 2020), we provide an effective approach for robust monitoring of highly-dynamic limb motions using a homemade wearable device with tailored algorithm. We demonstrate accurate and real-time measurements on three-dimensional (3D) motion velocities, accelerations and attitude angles of limbs in human daily activities, strenuous and prolonged exercises.

In our research, we propose a wearable motion capture device capable of implementing accurate and robust motion capture of human limb. We developed a 3D velocity micro sensor based on awareness of motion-induced surface flow vectors using micro flow sensors and combined it with tri-axis micro inertial sensors to constitute a wearable device. The developed device is competent to accurately measure 3D velocity, acceleration, angular velocity, and attitude angles of limb in dynamic motions. We also propose an integral-free data fusion algorithm to determine motion velocity, acceleration, and attitude angles to avoid accumulative errors and thus overcomes drift and instability problems faced by conventional inertial methods. In addition, we verify a natural intra-limb coordination relationship exists between shank and thigh in human walking and running, and establish a neural network model to represent this nature coordination relation. Using the intra-limb coordination model, thigh motion can be determined from shank motion, thereby dynamic motion capture of human lower limb including thigh and shank can be tactfully implemented by single shank-wearing device. This configuration greatly simplifies the motion capture system, and reduces the cost and alignment complexity of wearable device.

We demonstrate effective motion captures in boxing and kicking activities of Chinese Kungfu and long-time running, which validate excellent performance and robustness of the developed wearable device in capturing 3D dynamic motion. The developed device is used to monitor and evaluate physical states of full energy and fatigue. And we also apply the device to monitor patient suffering from mild meniscus injury, and show a potential indicator of knee flexion ability due to pain or pathological knee constraints. These demonstrations witness the promising potential of the developed device for applications in senior care, physical rehabilitation, daily life-logging and personal fitness.

In conclusion, this work shows a wearable motion capture device for accurately monitoring 3D limb motion in human dynamic activities. We hope that our work paves the way for wearable applications in human motor function assessment to ease healthcare burden, and also promotes wearable exoskeleton to improve human-robot interaction, etc.

The paper:    

S. Liu, J. Zhang, Y. Zhang, R. Zhu*, “A wearable motion capture device able to detect dynamic motion of human limbs”, Nature Communications, 11, 5615, 2020.