Step Detection Using Acceleration Variance Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional personal navigators inaccurately detect steps in environments without open sky GPS, as they fail to distinguish between mark-time and walking steps, leading to incorrect step length estimation and navigation solutions.
Innovation Solution
A step detecting apparatus and method using a multi-axis accelerometer sensor to calculate and compare the variance of acceleration signals in two directions, determining whether a step is a mark-time or walking step by comparing the variance sum with a threshold, allowing for accurate counting of both types of steps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If step detection is performed using only accelerometer output patterns, then the detection process is simple, but mark-time steps are incorrectly identified as walking steps leading to inaccurate step length estimation
Solution Approach 1:
The patent changes the parameter used for step detection from simple accelerometer output patterns to the variance of acceleration signals. By calculating variance in multiple directions and comparing it to a threshold, the system can distinguish between mark-time steps (low variance) and walking steps (high variance), thereby improving detection accuracy while maintaining relatively simple processing
Solution Approach 2:
The patent introduces a new dimension of analysis by calculating variance in multiple directions (X, Y, Z axes) rather than relying on a single accelerometer output pattern. This multi-dimensional variance analysis enables the system to differentiate between different types of steps more accurately
2Measurement precision
If GPS information is used for navigation, then navigation accuracy is high in open sky environments, but GPS signals are unavailable indoors or in city centers
Solution Approach 1:
The patent makes the navigation system universal by enabling it to function in multiple environments (open sky and indoor/city center) using different methods. The system uses GPS when available and switches to step-based navigation when GPS is unavailable, achieving multi-functional adaptability across different operational contexts
3Device complexity
If step length estimation is performed without distinguishing mark-time and walking steps, then the processing is straightforward, but the navigation solution becomes incorrect
Solution Approach 1:
The patent changes the parameter for step classification from simple step detection to variance-based classification. By computing variance of acceleration signals in multiple directions and comparing against a threshold, the system reliably distinguishes between mark-time steps and walking steps, ensuring accurate step length estimation and reliable navigation solutions
Data Source
AI summary
A step detecting apparatus and method in a personal navigator where acceleration signals in at least two directions are acquired from an accelerometer sensor, as a pedestrian moves. Variances of the acceleration signals are calculated and summed. The variance sum is compared with a threshold. If the variance sum is greater than the threshold, it is determined that a detected step is a walking step. If the variance sum is less than the threshold, it is determined that the detected step is a mark-time step.