Biaxial Sensor Walking Detection via Frequency Ratio Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for detecting walking activity in individuals face challenges in reliability and accuracy, particularly in setting universal thresholds for peak amplitude detection and distinguishing walking from other activities, leading to inconsistent results.
Innovation Solution
A system utilizing a biaxial or triaxial movement sensor attached to the body, processing signals over a time window to find a dominant frequency ratio between axes, with high-pass and pass-band filters to enhance precision and reduce noise, allowing for automatic and robust detection of walking activity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If a universal threshold is set a priori for peak amplitude detection, then the system can operate automatically, but the reliability of detection deteriorates due to inability to adapt to different individuals and conditions
Solution Approach 1:
The patent changes the detection parameter from fixed amplitude threshold to frequency ratio. The system calculates the ratio between dominant frequency and fundamental frequency, which remains relatively stable across different individuals and walking conditions. This parameter transformation enables both automatic operation and reliable detection without requiring subject-specific calibration.
2Device complexity
If simple peak amplitude thresholding is used, then the device complexity is reduced, but the ability to distinguish walking from other activities deteriorates
Solution Approach 1:
The patent transforms the detection approach by changing from amplitude-based to frequency-based parameters. By analyzing the ratio between dominant frequency and fundamental frequency, the system can distinguish walking patterns from other activities more effectively while maintaining computational simplicity.
Solution Approach 2:
The patent moves the detection from one-dimensional amplitude analysis to two-dimensional frequency domain analysis. By examining both fundamental frequency and dominant frequency components, the system gains additional discriminatory power for distinguishing walking from non-walking activities.
3Speed
If frequency analysis is performed without time windowing, then the processing speed is improved, but the precision of walking detection deteriorates due to non-stationary signal characteristics
Solution Approach 1:
The patent segments the continuous acceleration signal into fixed time windows (e.g., 10-second intervals). Within each window, the signal is treated as stationary and frequency analysis is performed. This segmentation approach balances processing speed with detection precision by analyzing manageable signal segments rather than the entire continuous signal.
Solution Approach 2:
The patent performs preliminary filtering and segmentation of the signal before frequency analysis. By pre-processing the signal into standardized time windows and applying appropriate filters, the system prepares the data for efficient frequency domain transformation, optimizing both speed and precision.
Data Source
AI summary
A system for detecting the walk of a person has a housing (BT) with at least a biaxial movement sensor (CM). The housing is attached to the upper portion of the body of the person, so that a first measurement axis of the sensor (CM) coincides with the anteroposterior axis (AP) or the vertical axis (VT) of the body and so a second measurement axis of said sensor (CM) coincides with the mediolateral axis (ML) of the body. An analysis means (MA) analyzes the measurements delivered by the sensor (CM). The analysis means includes a processing means (MT) for processing over a time window the measurement signals delivered by the sensor (CM), which includes means for searching for a dominant frequency (MRFD) in said signals. The analysis means also includes a detection means (MD) for detecting the walk of the person when a ratio between the dominant frequency of the signal of the first measurement axis and the dominant frequency of the second measurement axis, or between the dominant frequency of a Euclidian norm of the vector of measurements transmitted by the sensor (CM) and the dominant frequency of the signal of the second measurement axis, is substantially equal to two.