Dynamic Accelerometer Thresholds for Step Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing step-counting systems using accelerometers face challenges in accurately detecting steps due to varying acceleration patterns caused by factors like walking speed, surface elasticity, and shoe hardness, leading to inaccuracies with fixed thresholds, particularly for individuals with different activity levels such as healthy individuals and post-surgical patients.
Innovation Solution
The system employs dynamic sequential thresholds that adjust based on sensed accelerometer signals, using a rest threshold, peak threshold, and amplitude threshold to differentiate between step signals and noise, allowing for accurate step counting across varying intensities and conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed acceleration threshold is used to filter noise, then the device complexity is reduced, but the measurement precision deteriorates because the threshold cannot adapt to varying step intensities across different individuals and conditions
Solution Approach 1:
The patent implements dynamic threshold adjustment by continuously adapting the acceleration threshold based on the statistical properties (mean and standard deviation) of recent accelerometer readings. This allows the threshold to automatically adjust to varying step intensities across different individuals, walking speeds, and surface conditions, resolving the contradiction between simple filtering and accurate detection
Solution Approach 2:
The system changes the threshold parameter dynamically by calculating it as a function of the mean and standard deviation of recent readings. This parameter adaptation enables the threshold to scale with the user's typical step intensity, maintaining measurement precision without requiring complex device architecture
2Reliability
If a high acceleration threshold is set to avoid false positives from noise, then the reliability of noise filtering improves, but the measurement precision deteriorates because genuine steps from post-surgical patients or light walkers are missed
Solution Approach 1:
The threshold parameter is dynamically adjusted based on the user's historical step data, calculating the mean and standard deviation of recent readings. This allows the system to set appropriate thresholds for each user's typical step intensity, maintaining high reliability for their specific walking pattern while avoiding false negatives
Solution Approach 2:
The system uses feedback from accumulated accelerometer readings to continuously refine the threshold setting. By monitoring the distribution of recent readings and adjusting the threshold based on statistical properties, the system maintains reliable noise filtering while adapting to changes in the user's walking pattern or condition
3Measurement precision
If a low acceleration threshold is set to capture all genuine steps, then the measurement precision improves for light walkers, but the reliability deteriorates because more false positives from noise are counted as steps
Solution Approach 1:
The threshold is dynamically adjusted to match the user's typical step intensity profile. By calculating the mean and standard deviation of recent readings, the system automatically sets a threshold that is sensitive enough to detect genuine steps while remaining high enough to filter noise, adapting to each user's characteristics rather than using a fixed low threshold
4Adaptability or versatility
If multiple dynamic thresholds are implemented to adapt to varying conditions, then the adaptability improves for different user conditions, but the device complexity increases due to additional processing requirements
Solution Approach 1:
The system achieves adaptability through parameter changes rather than structural complexity. By dynamically adjusting the threshold parameter based on simple statistical calculations (mean and standard deviation of recent readings), the system adapts to different users and conditions without requiring complex algorithms or additional hardware
Solution Approach 2:
The threshold adjustment mechanism is self-service, automatically adapting to each user's walking pattern without requiring manual calibration or complex external processing. The system uses its own accelerometer data to compute and adjust thresholds, minimizing additional processing requirements while maximizing adaptability
Data Source
AI summary
Systems and methods detect steps from one or more sensed accelerometer signals. Systems comprise an accelerometer and a non-transitory computer readable medium, each of which communicates with a processor. The accelerometer is coupled to an individual and generates outputs received by the processor. The non-transitory computer readable medium stores instructions controlling the processor to perform steps of a method. The processor determines a minimum and a maximum reading for respective time periods. Counted peak heights are maxima that exceed a rest threshold. The rest threshold may be a function of the rest maximum. The processor determines a peak threshold for each time period having a counted peak height. A counted amplitude comprises the difference between the maximum reading and minimum readings of a time period in which the counted peak height exceeds the peak threshold. The processor increments a step counter when a counted amplitude exceeds an amplitude threshold.