Comparison of step length and heading estimation methods for indoor environments

Abstract. With the continuous improvement of the hardware level of the inertial measurement unit (IMU), indoor pedestrian dead reckoning (PDR) using an inertial device has been paid more and more attention. Typical PDR system position estimation is based on acceleration obtained from accelerometers to measure the step count, estimate step length and generate the position with the heading received from angular sensors (magnetometers and gyroscopes). Unfortunately, collected signals are very responsive to the alignment of sensor devices, built-in instrumental errors and distortions from the surrounding environment. In our work, a pedestrian positioning method using step detection based on a shoe-mounted inertial unit is arranged and put to the test, and the final results are analyzed. The extended Kalman filter (EKF) provides estimation of the errors which are acquired by the XSENS IMU sensor biases. The EKF is revised with acceleration and angular rate computations by the ZUPT (zero velocity update) and ZARU (zero angular rate update) algorithms. The step detection associated with these two solutions is the perfect choice to calculate the current position and distance walked and to estimate the IMU sensors' collected errors by using EKF. The test with a shoe-mounted IMU device was performed and analyzed in order to check the performance of the recommended method. The combined PDR final results were compared to GPS/Beidou postprocessing kinematic results (outdoor environment) and to a real route which was prepared and calculated for an indoor environment. After the comparison, the results show that the accuracy of the regular-speed walking under ZUPT and ZARU combination in the case of outdoor positioning did not go beyond 0.19 m (SD) and for indoor positioning accuracy did not exceed 0.22 m (SD). The authors are conscious that built-in drift errors coming from accelerometers and gyroscopes, as well as the final position obtained by XSENS IMU, are only stable for a short time period. Based on this consideration, our future work will be focused on supporting the methods presented with radio technologies (WiFi) or image-based solutions to correct all IMU imperfections.

The demand for pedestrian navigation has increased along with the rapid progress in mobile and wearable devices. This study develops an accurate and usable Step Length Estimation (SLE) method for a Pedestrian Dead Reckoning (PDR) system with features including a wide range of step lengths, a self-contained system, and real-time computing, based on the multi-sensor fusion and Fuzzy Logic (FL) algorithms. The wide-range SLE developed in this study was achieved by using a knowledge-based method to model the walking patterns of the user. The input variables of the FL are step strength and frequency, and the output is the estimated step length. Moreover, a waist-mounted sensor module has been developed using low-cost inertial sensors. Since low-cost sensors suffer from various errors, a calibration procedure has been utilized to improve accuracy. The proposed PDR scheme in this study demonstrates its ability to be implemented on waist-mounted devices in real time and is suitable for the indoor and outdoor environments considered in this study without the need for map information or any pre-installed infrastructure. The experiment results show that the maximum distance error was within 1.2% of 116.51 m in an indoor environment and was 1.78% of 385.2 m in an outdoor environment.

Purpose Step length is a key factor for pedestrian dead reckoning (PDR), which affects positioning accuracy and reliability. Traditional methods are difficult to handle step length estimation of dynamic gait, which have larger error and are not adapted to real walking. This paper aims to propose a step length estimation method based on frequency domain feature analysis and gait recognition for PDR, which considers the effects of real-time gait. Design/methodology/approach The new step length estimation method transformed the acceleration of pedestrians from time domain to frequency domain, and gait characteristics of pedestrians were obtained and matched with different walking speeds. Findings Many experiments are conducted and compared with Weinberg and Kim models, and the results show that the average errors of the new method were improved by about 2 meters to 5 meters. It also shows that the proposed method has strong stability and device robustness and meets the accuracy requirements of positioning. Originality/value A sliding window strategy used in fast Fourier transform is proposed to implement frequency domain analysis of the acceleration, and a fast adaptive gait recognition mechanism is proposed to identify gait of pedestrians.

In this paper, we develop a foot‐mounted pedestrian navigation system prototype with the emphasis on distance measuring with an inertial measurement unit (IMU) which implies the characteristics of pedestrian gait cycle and thus can be used as a crucial step indicator for distance calculation. Conventional methods for step detection and step length estimation cannot adapt well to the general pedestrian applications since the parameters in these methods may vary for different persons and motions. In this paper, an adaptive time‐ and frequency‐domains joint distance measuring method is proposed by utilizing the means of behaviors classification. Two key issues are studied: step detection and step length determination. For the step detection part, first behavior classification along with state transition strategy is designed to identify typical pedestrian behaviors including standing still, walking, running and irregular swing. Then a four‐stage step detection method is proposed to adaptively determine both step frequency and threshold in a flexible window. Based on the behavior classification results, a two‐segment functional based step length model is established to adapt the walking and running behaviors. Finally, real experiments are carried out to verify our proposed step detection method and step length model. The results show that the proposed method outperforms the existing representative methods and it exhibits the merits of accuracy and adaptability for different persons in real time and significantly improves the accuracy of distance measuring.

The effects of electrical vestibular stimulation upon movement and perception suggest two evoked sensations: head roll and inter-aural linear acceleration. The head roll vector causes walking subjects to turn in a direction dependent on head pitch, requiring generation of torque around a vertical axis. Here the effect of vestibular stimulation upon vertical torque (T(z)) was investigated during quiet stance. With the head tilted forward, square-wave stimuli applied to the mastoid processes evoked a polarity-specific T(z) response accompanied by trunk yaw. Stochastic vestibular stimulation (SVS) was used to investigate the effect of head pitch with greater precision; the SVS–T(z) cross-correlation displayed a modulation pattern consistent with the head roll vector and this was also reflected by changes in coherence at 2–3 Hz. However, a separate response at 7–8 Hz was unaffected by head pitch. Head translation (rather than rotation) had no effect upon this high frequency response either, suggesting it is not caused by a sense of body rotation induced by an inter-aural acceleration vector offset from the body. Instead, high coherence between medio-lateral shear force and T(z) at the same frequency range suggests it is caused by mechanical coupling to evoked medio-lateral sway. Consistent with this explanation, the 7–8 Hz response was attenuated by 90 deg head roll or yaw, both of which uncouple the inter-aural axis from the medio-lateral sway axis. These results demonstrate two vertical torque responses to electrical vestibular stimulation in standing subjects. The high frequency response can be attributed to mechanical coupling to evoked medio-lateral sway. The low frequency response is consistent with a reaction to a sensation of head roll, and provides a novel method for investigating proprioceptive-vestibular interactions during stance.

The use of adaptive physical culture in the system of complex rehabilitation of patients with cerebral palsy (CP) is becoming more popular. Adaptive climbing (AC) is one of these types of physical culture and sports activities. Currently, there are not many data in the domestic and world literature devoted to the study of the AC use in the complex rehabilitation of patients with CP. Aim. Evaluation of the effectiveness and safety of AC in the rehabilitation of children and adolescents with spastic forms of CP in addition to conventional medical rehabilitation. Material and methods. The study included 30 patients aged from 5 to 18 years with spastic forms of CP who were allocated into one of two groups: the 1st group, the study group (n=15) — patients who underwent standard medical rehabilitation and AC classes; the 2nd group, the control group (n=15) — participants who received only medical rehabilitation. A comprehensive assessment of the rehabilitation measures effectiveness and safety was performed before the start of AC classes and after 3 months using clinical neurological and instrumental examinations, including biomechanical examination (electropodography, video analysis of walking and stabilography). Results. Group 1 patients showed a statistically significant improvement in general motor functions on the GMFM-88 scale (p=0,0001), a decrease in the level of spasticity on the мodified Ashworth scale (p=0,0051) and an improvement in manual abilities on the MACS scale (p=0,0431) 3 months after inclusion in the study. Positive dynamics in the form of an increase in the symmetry of step time (p=0,0063), step length (p=0,0083) and step velocity (p=0,0409) were noted according to the electropodographic study in the 1st group. A statistically significant increase in the range of motion in the hip (p=0,0038) and knee joints (p=0,0076) was detected in the 1st group according to video analysis of walking. There was no significant change in the indicators in the 2nd group, except for the severity of spasticity on the modified Ashworth scale (p=0,0180). A positive correlation was found between the indicators “step length” and “ step velocity” (r=0,605, p=0,0168). A negative correlation was found between the “step velocity “ indicator and the GMFCS score (r= – 0.832, p=0,0372). Only 2 adverse events were registered in the study group: acute respiratory viral infection and wrist injury during AC. Conclusion. AC, in addition to traditional methods of medical rehabilitation, leads to a decrease in spasticity, increased joint mobility and control of voluntary movements, improved maintenance of the body position, walking and precise hand movements, and also has a favorable safety profile. AC is an effective and safe motivating method of physical culture that can change the patient’s view on traditional therapy.

Indoor Location-Based Service (ILBS) shows great research promotions with wide applications e.g., indoor firefighting and cave exploration. Foot-mounted Inertial Navigation System (INS), one approach of ILBS, lacks a reference map of the environment, resulting in poor trajectory recognition. This paper introduces SLAMING, a novel wearable type SLAM via a Zero Angular rate Update (ZARU) aided Inertial NaviGation. SLAMING proposes a gravity center calculation method, merging the dual (left and right) foot trajectories. Moreover, the proposed polar projection and occupancy grid map method determines the map boundary, enabling the fusion of the trajectory and ultrasound range. The mapping results of SLAMING are demonstrated with the ground truth. The location performance is validated using a self-created database, the results of which indicate lower horizontal and spherical error compared with the traditional INS in all scenarios.

In a waist-worn Pedestrian Navigation System (PNS) based on Dead-Reckoning (DR), heading drift caused by Micro-Electro-Mechanical System (MEMS) gyro bias is an essential factor affecting DR accuracy. Considering the characteristics of pedestrian navigation and the poor bias repeatability of MEMS gyros, this paper presents a standing calibration method for MEMS gyro bias. The current gyro biases for each boot can be calibrated accurately in the initial stage before walking. Since the attitude angles calculated by the output data from magnetic sensor and accelerometers do not drift, this paper applies the reverse algorithm of attitude updating to calculate the angular velocities of human motion. Then the gyro biases at each moment can be acquired by subtraction operation between the measured angular velocities from gyros and the calculated angular velocities of human motion. Finally, in order to restrain the random error caused by sensor noise, the calculated biases in the initial stage are smoothed, and consequently the optimal estimate of current gyro biases after each boot can be obtained. Still and dynamic turntable experiments and a walking experiment are performed to compare and analyse the proposed method and the Zero Angular Rate Update (ZARU) method. Experimental results show that the proposed method can also calibrate the gyro bias accurately in the case of body swaying.

The third generation Intelligent Device for Energy Expenditure and Activity (IDEEA3, MiniSun, CA) has been developed for clinical gait evaluation, and this study was designed to evaluate the accuracy and reliability of IDEEA3 for the gait measurement of lumbar spinal stenosis (LSS) patients. Twelve healthy volunteers were recruited to compare gait cycle, cadence, step length, velocity, and number of steps between a motion analysis system and a high-speed video camera. Twenty hospitalized LSS patients were recruited for the comparison of the five parameters between the IDEEA3 and GoPro camera. Paired t-test, intraclass correlation coefficient, concordance correlation coefficient, and Bland-Altman plots were used for the data analysis. The ratios of GoPro camera results to motion analysis system results, and the ratios of IDEEA3 results to GoPro camera results were all around 1.00. All P-values of paired t-tests for gait cycle, cadence, step length, and velocity were greater than 0.05, while all the ICC and CCC results were above 0.950 with P < 0.001. The measurements for gait cycle, cadence, step length, velocity, and number of steps with the GoPro camera are highly consistent with the measurements with the motion analysis system. The measurements for IDEEA3 are consistent with those for the GoPro camera. IDEEA3 can be effectively used in the gait measurement of LSS patients.

Personal Dead Reckoning based on foot-mounted Inertial Measurement Units is a research hotspot in the field of positioning and navigation in recent years. This paper conducts a targeted research on the application of current mainstream attitude and heading reference system (AHRS) algorithm in the foot inertial navigation positioning. Through open datasets, the positioning accuracy and directional accuracy of 9-state complementary Kalman filter (CKF) are compared and analyzed among the conventional algorithm, Mahony algorithm, and Madgwick algorithm, in which the Madgwick algorithm can achieve the best positioning results. And on this basis, for the Madgwick algorithm, it is verified that it can help improve the positioning accuracy of 15-state CKF under the assistive technologies of zero angular rate update (ZARU) and heuristic heading reduction (HDR). The adaptive zerospeed detection algorithm is designed, and the threshold value of zero-speed detection is set dynamically through tracking the variable of speed in CKF, which can detect the time period of zero-speed state more accurately, thus further improving the correction of directional errors. Finally, the effectiveness of the proposed algorithm is further proved by actual data.

Gait impairments are a well-documented consequence of stroke, yet the impact of transient ischemic attack (TIA) on gait remains underexplored. This study aimed to characterize acute changes in gait patterns within three days of stroke or TIA onset using an accelerometer, and to examine the relationship between gait patterns and questionnaire-based measures of balance confidence. We found that patients with TIA showed significantly lower variability in swing time, stance time, and step time than patients with stroke. Comparisons between patients with TIA and controls indicated that controls had higher mean step velocity and lower step time asymmetry. Additionally, compared with controls, patients with stroke demonstrated a slower and less temporally consistent pace, indicated by shorter step length, lower step velocity, and higher standard deviation (SD) for swing time, stance time, and step time. They also showed greater variability, reflected by higher step length SD, and more pronounced gait asymmetry, with greater asymmetry in swing time, step time, and stance time. The balance confidence was associated with gait parameters in patients with stroke and in controls, but not in patients with TIA. In both groups, higher confidence related to longer step length and faster step velocity; in stroke it was also linked to lower swing time variability and longer step and stance times, and in controls to lower step time variability. These findings demonstrate that TIA involves detectable neurological changes affecting motor control, highlighting the value of detailed accelerometer-based gait analysis for early detection of functional deficits.

PurposeThe study aim was to analyse the relationships between the results of the Timed Up and Go (TUG) test and the Functional Reach Test (FRT), and the temporal and spatial gait parameters determined with the GAITRite system.MethodsThe study included 60 healthy, physically active elderly people aged 70.4 ± 5.1 years. The participants’ functional fitness was evaluated with the TUG and FRT, and their temporal and spatial gait parameters – with the GAITRite system.ResultsThe TUG results correlated inversely with the step length (&lt;i&gt;r&lt;/i&gt; = –0.70 and &lt;i&gt;r&lt;/i&gt; = –0.61 for the preferred and fast speed, respectively), stride length (&lt;i&gt;r&lt;/i&gt; = –0.71 and &lt;i&gt;r&lt;/i&gt; = –0.61, respectively), and velocity (&lt;i&gt;r&lt;/i&gt; = –0.69 and &lt;i&gt;r&lt;/i&gt; = –0.38, respectively). The regression model with these variables explained 43% of variance in the TUG results (p &amp;#8804 0.0001). The FRT results correlated positively with the step length (&lt;i&gt;r&lt;/i&gt; = 0.61 and &lt;i&gt;r&lt;/i&gt; = 0.54, for the preferred and fast speed, respectively), stride length (&lt;i&gt;r&lt;/i&gt; = 0.60 and &lt;i&gt;r&lt;/i&gt; = 0.56, respectively), and velocity (&lt;i&gt;r&lt;/i&gt; = 0.37 and r = 0.38). The regression model with these variables explained 26% of variance in the FRT outcomes.ConclusionsBasic temporal and spatial parameters of gait at the preferred speed, i.e. velocity, step length, stance time, single support time, swing time, and double support time, explain up to 43% of the TUG outcome variance and 26% of the FRT results variance.

Crouch gait in persons with positive sagittal spine alignment resolves with surgery

The present study was undertaken in effort to describe the relationships among each step length, stride length, stride frequency, and velocity during the start dash. Stride frequency reached its peak immediately after the start, while stride length required about 25-30 strides to reach its maximal level. Although substantial increase in velocity was attained by an increase in stride length, it appears that acceleration during the start dash is achieved by maximal increase in stride frequency. Mid step length reached a maximum level following several strides, however airborne step length required more time to reach its maximum. Therefore acceleration during the start dash is achieved by a maximal increase in mid step length since it is not possible to quickly increase airborne step length.

Purpose/Hypothesis: Virtual Reality (VR) has been proposed as a strategy for rehabilitation. To measure the potential of new VR devices they need to be characterized and compared to their real world counterparts. The primary purpose of this study was to characterize and validate the gait of individuals who walked on a VR walking simulator (WS) by comparing it to the kinematics and velocity of overground walking. The secondary purpose was to determine if the presence of the virtual environment (VE) modified walking characteristics. We hypothesized that walking velocity would be greater overground and slowest without the VE. Number of Subjects: Ten healthy subjects (2 male and 8 female) ages 24–53 without musculoskeletal injuries or neuromuscular pathology participated. Materials/Methods: Gait step length and velocity were collected (three trials) as participants walked at their self selected speed on a Gait Rite mat. Participants wore a harness and refective markers on the foot, ankle, knee, hip and trunk. They stood on the WS with each foot secured on a robotic platform, a harness attached to an unweighing frame, facing a screen. Unweighed 40% of their body weight they practiced walking on the WS (2–8 min.) until achieving the criterion gait pattern. Data were collected for four street crossing trials, walk without VE (2), walk with VE (2). An auditory tone or the VE green light prompted participants to begin walking. Kinematics were acquired using a 6 camera Peak system at 60 Hz. Joint angles were calculated for push of and initial contact by averaging the peak angle of three steps. Velocity data were averaged from the x-translation plots of the ffth metatarsal. Joint excursions from push of to initial contact-loading response were compared to normative data. Differences between overgound walking velocity and step length relative to the mobility simulator with and without the VE were assessed using a repeated measures ANOVA (alpha level= .05). Results: Walking velocity and step length were significantly greater for overground walking (1.5 m/sec; .79 m) compared with the mobility simulator with (.22 m/sec; .17 m) and without (.21 m/s; .18 m) the VE. The average joint excursions on the mobility simulator for push of to initial contact-loading response were 34 (knee) and 18 (ankle) degrees; compared to a range from push of to initial contact-loading response of 22–40 (knee) and 10–20 (ankle) degrees in normal gait. Conclusions: Walking in the VE resulted in pronounced decreases in temporal distance measures of gait however joint excursions were comparable to normal gait. There were no differences in gait velocity, step length or joint excursions when walking on the mobility simulator with and without the VE. The mechanics of the robot limit step length however, changing street crossing time in the VE may increase walking velocity. Clinical Relevance: Selected aspects of gait were preserved in the WS while others were not. Transfer of training from the walking simulator to real world walking still needs to be determined.