Accelerometer Peak Detection with Non-Uniform Sampling Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for automatically identifying modes of transportation using accelerometer data require long time windows, which is impractical due to limited storage capacity in embedded systems, and existing approaches degrade performance when trying to compress data.
Innovation Solution
A method that involves non-uniform sampling of accelerometer measurements to retain local extrema and resample based on time and amplitude thresholds, allowing for efficient detection and characterization of acceleration peaks without degrading recognition performance, including calculating characteristics like average amplitude, duration, and frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If regular uniform sampling is used to capture acceleration data, then measurement precision is maintained, but data volume increases requiring more storage capacity
Solution Approach 1:
The patent extracts only the essential information from acceleration data by identifying and retaining local extrema (peaks and valleys) while discarding redundant intermediate samples. This extraction principle maintains the critical characteristics needed for transportation mode detection while significantly reducing data volume stored in buffer memory.
Solution Approach 2:
The patent segments the continuous acceleration signal into discrete events (acceleration peaks) by detecting local extrema. Each peak is characterized by specific parameters (amplitude, duration, timing) rather than storing the entire continuous signal, enabling efficient storage and processing of transportation mode information.
2Quantity of substance
If data compression is applied to reduce storage requirements, then storage capacity is optimized, but recognition performance degrades
Solution Approach 1:
The patent applies different processing quality to different parts of the acceleration signal. Local extrema points are preserved with high fidelity while intermediate points are discarded. This local quality approach ensures that the most informative parts of the signal (the peaks indicating acceleration events) are maintained while reducing overall data volume.
Solution Approach 2:
The patent transforms the acceleration signal from time-domain continuous values to a set of discrete parameters characterizing each peak (amplitude, duration, timing relative to other peaks). This parameter transformation maintains the essential information needed for transportation mode classification while enabling efficient storage and processing.
3Measurement precision
If long time windows are used to distinguish between transportation modes, then measurement precision improves, but loss of time increases
Solution Approach 1:
The patent performs preliminary processing of acceleration data by continuously identifying and characterizing peaks as they occur. By maintaining a buffer of recent peaks and their characteristics, the system is prepared to quickly classify transportation modes as new data arrives, reducing the effective processing time needed while maintaining accuracy.
Solution Approach 2:
The patent uses the periodic occurrence of acceleration peaks as natural sampling points for analysis. Rather than continuously analyzing all data points, the system periodically evaluates peak characteristics (amplitude, duration, timing patterns) to determine transportation mode, enabling efficient real-time classification.
Data Source
Figure 1
Figure 2a~2b
Figure 3
AI summary
The invention relates to a method for determining information about the displacement of an object from measurements provided by an accelerometer attached to the object. The method comprises the steps of detecting acceleration peaks in the measurements, calculating one or more characteristics of the detected acceleration peaks, and determining a displacement pattern of the object from the characteristics of the acceleration peaks. The measurements are subjected to non-uniform resampling based on the detection of local extrema.