Motion Signal Feature Counting via Dynamic Reference Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wearable devices face challenges in accurately counting activity cycles due to noise and artifacts in motion signals, particularly in distinguishing between regular and irregular peaks, which affects precision in monitoring physical activities.
Innovation Solution
A system and method for counting features in motion signals using a sensor and controller to detect valid features by determining periods, locations, and values within the signal, with a reference function to validate and filter out noise, allowing for precise counting of activity cycles in wearable devices and other applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If low pass and high pass filters are used to remove noise and artifacts, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the signal processing task into distinct phases: detecting positive peaks, detecting negative peaks, determining periods, and validating features. This segmentation allows each processing step to be optimized independently, improving measurement precision without proportionally increasing overall system complexity
Solution Approach 2:
The patent performs preliminary actions by pre-determining the level threshold and reference function before feature validation. The reference function is established in advance based on signal characteristics, enabling faster real-time validation without adding complex real-time computation
2Reliability
If artifact peaks are filtered out using traditional filtering methods, then reliability is improved, but manufacturing precision deteriorates due to difficulty in distinguishing artifact peaks from valid peaks
Solution Approach 1:
The patent introduces a reference function as an intermediary between the raw signal and the decision to count a peak. This reference function, derived from signal characteristics, mediates the validation process by providing a dynamic threshold that distinguishes valid peaks from artifacts more reliably than fixed filtering methods
Solution Approach 2:
The patent implements dynamic validation by adjusting the reference function based on signal characteristics. The validation threshold is not fixed but adapts to the actual signal conditions, allowing the system to maintain high reliability across varying activity types and conditions
Data Source
AI summary
Systems and methods for counting features in a motion signal are provided. An example method includes detecting a signal indicative of a motion of a human body, determining, by a controller communicatively coupled to the sensor, a period in the signal, determining a location and a value of a feature within the period of the signal, determining that the location and the value of the feature are valid for counting, and in response to the determination that the location and the value of the feature are valid for counting, increasing a counter for the feature. The feature may also include a global maximum of the signal within a positive phase of the period. Criteria for the validation depend on a current type of activity. The method includes automatically determining the current type of activity based on several preceding locations and values of the feature in the signal.