A motion capture algorithm based on inertia-Kinect sensors for lower body elements and step length estimation. (February 2021)

• Record Type: Journal Article
• Authors: Abbasi, Javad; Salarieh, Hassan; Alasty, Aria
• Is Part Of: Biomedical signal processing and control. Volume 64(2021)
• Publication Date: 2021-02
• Subjects: Motion capture -- Data fusion -- Gait analysis -- Step length estimation -- Kinect -- Inertial sensor; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2020.102290
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: An integrated IMU-Magnetometer-Kinect motion capture algorithm is proposed. It is based on Unscented Kalman Filter and also an optimization based method. It estimates the walking step length while compensating magnetometer distortions. It can cope with biases and accumulated errors of IMU and occlusion error of Kinect. A great improvement is achieved in joint position estimation comparing to Kinect alone. Abstract: Motion capture is a process that movements of living organisms like human or objects are captured and the results are processed for the desired applications. These applications are in rehabilitation, sports, film industry and etc. There are many techniques and instruments for motion capture that optical camera systems are the most accurate ones. But these cameras are high cost and limited to labs. Some sensors like Inertial Measurement Units (IMU) and recently, Kinect cameras have been considered by many researchers because these are low cost and easy to use. But problems like bias, accumulated error and occlusion make them look for improvements. Fusion algorithms are one of the best methods that help to use from each sensor's strengths. The purpose of this work is design and implementation of an efficient algorithm for estimation of lower limbs joints 3D positions and step length. Orientation quaternions are considered as estimation states. An algorithm was developed with gradient descent and unscented Kalman filter approach based on IMUs and Kinect's measurements. The IMUs' data consist of three mutually orthogonal gyroscopes, three mutually orthogonal accelerometers, and a three-axis magnetometer. In this algorithm bias and magnetic distortions have been compensated in parallel structure. The resulted errors have been reported with respect to VICON optical camera system. The results obtained from an experimental test, show up to 60 percent improvement on Kinect in joints 3D positions estimation and the algorithm improves step length estimation error of Kinect from 7.8 cm to 0.03 cm.