Inertial Sensor-Based Step Length Estimation Model by Means of Principal Component Analysis

Abstract

Inertial sensor-based step length estimation has become increasingly important with the emergence of pedestrian-dead-reckoning-based (PDR-based) indoor positioning. So far, many refined step length estimation models have been proposed to overcome the inaccuracy in estimating distance walked. Both the kinematics associated with the human body during walking and actual step lengths are rarely used in their derivation. Our paper presents a new step length estimation model that utilizes acceleration magnitude. To the best of our knowledge, we are the first to employ principal component analysis (PCA) to characterize the experimental data for the derivation of the model. These data were collected from anatomical landmarks on the human body during walking using a highly accurate optical measurement system. We evaluated the performance of the proposed model for four typical smartphone positions for long-term human walking and obtained promising results: the proposed model outperformed all acceleration-based models selected for the comparison producing an overall mean absolute stride length estimation error of 6.44 cm. The proposed model was also least affected by walking speed and smartphone position among acceleration-based models and is unaffected by smartphone orientation. Therefore, the proposed model can be used in the PDR-based indoor positioning with an important advantage that no special care regarding orientation is needed in attaching the smartphone to a particular body segment. All the sensory data acquired by smartphones that we utilized for evaluation are publicly available and include more than 10 h of walking measurements.

Automatic Hajj and Umrah Ritual Detection Using IMU Sensors

“…is used to estimate the length of each step taken by the user [43], [44]. Weinberg formula is an empirical relation that is commonly utilized in PDR systems to estimate the step length with respect to the user's acceleration.…”

Adaptive Inertial Sensor-Based Step Length Estimation Model

“…In contrast, Weinberg [ 55 ] used the difference between maximum and minimum acceleration values within a step as the basis, whereas Kim et al [ 56 ] utilized the mean absolute acceleration value in the walking direction within the step. Vezočnik et al [ 57 ] based their model on the difference between maximum and minimum acceleration magnitude values within the step. Guo et al [ 58 ] and Bylemans et al [ 31 ] extended the model proposed by Kim et al [ 56 ].…”

“…In one of our previous studies, we conducted an in-depth analysis and comparison of representative step length estimation models and demonstrated that the performances of certain models change under different evaluation protocols when smartphone placements and walking speeds vary in trials [ 48 ]. Moreover, authors also rarely consider the kinematics of different body segments during walking and step lengths in the derivation process of the model [ 57 ]. Specifically, identifying similar characteristics related to the movement of anatomical landmarks by means of the kinematics of the human body during walking does not often lay the groundwork for a new model, as it did in our previous study in which we proposed a step length estimation model based on the difference between maximum and minimum acceleration magnitude values within the step [ 57 ].…”

“…Moreover, authors also rarely consider the kinematics of different body segments during walking and step lengths in the derivation process of the model [ 57 ]. Specifically, identifying similar characteristics related to the movement of anatomical landmarks by means of the kinematics of the human body during walking does not often lay the groundwork for a new model, as it did in our previous study in which we proposed a step length estimation model based on the difference between maximum and minimum acceleration magnitude values within the step [ 57 ]. In addition, we aimed to include the minimum number of input parameters in the model and compare the evaluation results of the proposed model against models that exhibit similar composition, i.e., step length is calculated as the product of a tunable constant and one input parameter.…”

“…For these reasons, we aimed to investigate the possibilities of extending the acceleration-based model we proposed in our previous work when considering the movement of anatomical landmarks during walking [ 57 ]. As the base of the model already includes acceleration magnitude input, we utilized canonical correlation analysis (CCA) to investigate the link between acceleration magnitude, step length, and other characteristics defining the movement of anatomical landmarks, resulting in the adaptive step length estimation model herein based on step frequency and acceleration magnitude inputs.…”