Step Detection False Alarm Reduction via Manipulation Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing step detection methods in inertial navigation systems, such as those used in smartphones and tablets, suffer from a high rate of false alarms due to device manipulations, which are difficult to differentiate from walking signals, leading to inaccurate step counting.
Innovation Solution
A method that detects steps by identifying events from motion sensors while distinguishing intervals of device manipulation, using bandpass filtering and adaptive thresholding of acceleration signals, and synchronizing step detection with manipulation detection to exclude false alarms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If simple thresholding of acceleration signal norm is used for step detection, then sensitivity is improved (low rate of missed detections), but false alarm rate increases (spurious detections during device manipulation)
Solution Approach 1:
The patent segments the acceleration signal into different frequency components using bandpass filtering (1-2 Hz for steps, 0.1-0.5 Hz for manipulation). This allows independent processing of step signals and manipulation signals, enabling the system to detect steps with high sensitivity while filtering out manipulation-induced false alarms through frequency-based separation.
Solution Approach 2:
The patent introduces an intermediary manipulation detection module that acts as a mediator between the step detection module and the final step count output. This intermediary module detects device manipulation events and provides feedback to suppress false alarm steps during manipulation intervals, thereby reducing false alarms while preserving step detection sensitivity.
2Reliability
If adaptive filtering of step frequency is applied to reduce false alarms, then false alarm rate decreases, but device complexity increases
Solution Approach 1:
The patent segments the processing into distinct functional blocks: step detection module, manipulation detection module, and step validation module. Each module performs a specific function with dedicated processing, avoiding the need for complex adaptive filtering while maintaining low false alarm rates through modular frequency-based separation and feedback control.
3Reliability
If manipulation detection is integrated with step detection, then false alarm reduction is improved, but processing time increases
Solution Approach 1:
The patent applies preliminary bandpass filtering to separate manipulation signals from step signals before final detection. By pre-filtering the acceleration signal into distinct frequency bands (1-2 Hz for steps, 0.1-0.5 Hz for manipulation), the system prepares the data in advance, enabling faster and more accurate detection without significant processing time penalty.
Solution Approach 2:
The patent uses periodic sampling and processing cycles where the manipulation detection module continuously monitors for manipulation events and provides periodic feedback to the step validation module. This periodic feedback mechanism efficiently suppresses false alarms without requiring continuous complex processing, maintaining low processing time while improving reliability.
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
The method effectively reduces false alarms and improves the accuracy of step detection by distinguishing between walking and device manipulation events, resulting in a more reliable step counting system.
Implementation Method 1
The invention relates to a step detector which may be used as a block PM in an inertial navigator... receiving at its input signals Acc(t) produced by an accelerometer
Data Source
AI summary
A step detection method comprises the following steps: a) based on at least one signal coming from at least one motion sensor rigidly attached to a portable device, detect events suitable to correspond to steps of a user carrying the portable device; b) based on at least one signal, or on at least one signal coming from another sensor also rigidly attached to the portable device, identify intervals of time during which the device is being manipulated; and c) identify as steps those, from amongst the events detected, which do not coincide with the intervals of time. A device for the implementing such a method, a mobile telephone and touchscreen tablet comprising such a device are also provided.