Accelerometer Data Frame Transformation for Step Stride Disambiguation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current devices and methods for measuring user movement, such as pedometers and activity monitors, face inaccuracies in distinguishing between step and stride periods due to reliance on threshold-based calculations, leading to errors in step count determination, which is critical for both leisure and medical applications where accuracy is paramount.
Innovation Solution
The solution involves transforming acceleration data from a device-mounted frame of reference to a user-centric frame of reference, including a direction of travel and a side-to-side direction, and using autocorrelation functions in these directions to disambiguate between step and stride periods, thereby improving the accuracy of movement analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If threshold-based calculations are used to distinguish step and stride periods, then the method is simple to implement, but measurement precision deteriorates leading to errors in step count determination
Solution Approach 1:
The patent changes the parameter used for analysis from simple threshold-based magnitude comparisons to autocorrelation functions that analyze temporal patterns. By computing autocorrelation at different lags and comparing peaks at expected step period lags versus stride period lags, the system achieves more accurate differentiation without complex hardware, resolving the contradiction between simplicity and precision.
Solution Approach 2:
The patent replaces the mechanical/threshold-based detection system with a signal processing approach using autocorrelation analysis. Instead of relying on fixed thresholds that fail to distinguish periodic motions, the system uses mathematical autocorrelation functions to identify temporal patterns, substituting a simpler but inaccurate mechanical approach with a more precise computational method.
2Device complexity
If existing thresholding methods are used, then device complexity is low, but measurement precision deteriorates in distinguishing step periods from stride periods
Solution Approach 1:
The patent replaces simple threshold-based mechanical detection with autocorrelation analysis. The system computes autocorrelation functions and identifies peaks at specific lags corresponding to step and stride periods, providing a more sophisticated yet computationally manageable approach that significantly improves measurement precision while keeping device complexity reasonable.
Solution Approach 2:
The patent transforms the analysis from binary threshold comparisons to continuous autocorrelation function evaluation. By analyzing the shape and position of peaks in the autocorrelation function at different lags, the system achieves superior differentiation accuracy without requiring excessive computational resources, thus resolving the contradiction between device complexity and measurement precision.
3Ease of operation
If accelerometer data is analyzed in the device frame of reference, then processing is straightforward, but measurement precision deteriorates due to inability to accurately determine user movement direction
Solution Approach 1:
The patent adds the dimension of orientation transformation to the analysis. By rotating the accelerometer data from the device's fixed frame of reference to a user-centric frame aligned with the direction of travel, the system gains the ability to accurately determine movement direction and distinguish between step and stride periods, while the transformation process remains computationally efficient.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the accuracy of step and stride period determination, reducing errors and providing more reliable movement data for both fitness tracking and medical applications, including clinical studies and athletic performance analysis.
Implementation Method 1
an accelerometer mounted in the user's device as the user is walking
Implementation Method 2
calculate the auto-correlation of this magnitude data over time periods
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
An apparatus and method are disclosed for determining movement information for a user that carries an accelerometer whilst moving. The apparatus receives acceleration data from the accelerometer that are defined relative to a frame of reference of the accelerometer. A transformation is determined and applied to the acceleration data or to data derived from the acceleration data to determine acceleration data is a frame of reference of the user that includes a direction of travel of the user and a side to side direction transverse to the direction of travel of the user. The acceleration data or data derived from the acceleration data is analysed to determine a time period corresponding either to a stride period or to a step period of the user as the user is walking or running; and information about accelerations in the side to side direction are used to disambiguate whether the determined time period corresponds to the stride period of the user or to the step period of the user.