Reliable orientation estimation for mobile motion capturing in medical rehabilitation sessions based on inertial measurement units

SenseHAR

“…HAR is of great importance because it encourages users to adopt a healthier lifestyle by increasing personal awareness about physical activities and its positive consequences on health. It also has a great significance in designing artificially intelligent human-computer interface [22] for various applications such as sign language [8] and medical rehabilitation [9]. While accelerometer is the most prominent sensor used in activity detection [6,63,69], gyroscope and magnetometer [3,41] are also integrated to improve the probability of correct predictions and to identify activities such as the mode of locomotion [27,34,55,70], posture [78], gait [33], fall detection [72], exercise [61] and sleep pattern [46].…”

“…These devices are equipped with multiple sensors [39] and Inertial measurement units (IMUs) are the most prominent sensors that have a combination of microelectromechanical (MEMS) accelerometers, gyroscopes, and magnetometers that enable the devices to infer the activities of a person. While, prior research identifies smartphones as the widely used sensing modality in HAR [44,55], smartwatches are also being utilized extensively to infer diverse gestures for various applications such as sign language [8] and medical rehabilitation [9]. Today most people own multiple smart devices such as smartphones, smartwatch, fitness bands, smart shoes, smart belts, etc.…”

Adaptive regulation of the weights of REQUEST used to magnetic and inertial measurement unit based on hidden Markov model

“…The work of attitude algorithms used by MIMU can be divided into three components, the first component is the information fusion of the outputs of a triad of gyroscope for one estimate of the attitude (named ComI), the second component is the information fusion of the outputs of a triad of acceleration and magnetometer for the other estimate of the attitude (named ComII, besides, this component can be divided further into two sub components, named ComIIA (roll and pitch angles estimation only) and ComIIB (azimuth angle estimation only), respectively), and the last component is the information re-fusion realised by the weighted sum of those two attitude estimates (named ComIII). The common-used attitude algorithms by MIMU are extended Kalman filter (EKF) and complementary filter (CF), and usually, the attitude estimation accuracy of EKF is higher than CF, while the consuming time of EKF is longer than CF [10,11]. A quaternion is usually taken as the representation of attitude for those two attitude algorithms.…”

Positioning and navigation of mobile robot with asynchronous fusion of binocular vision system and inertial navigation system

“…Traditional positioning methods commonly include the use of a global positioning system (GPS), 1 an odometer, 2 and/or an inertial measurement unit (IMU). 3 These key technologies can be found in autonomous navigation, 4 motion planning, 5 and simultaneous localization and mapping (SLAM). 6 As one of the popular localization methods, GPS is widely used for land vehicle positioning, although the signal propagated by a satellite may experience discontinuity and blockage caused by high-rise buildings or other obstacles.…”