Method and system for heterogeneous event detection
Inventors
Cheng, Chun Hing • PURDY, MICHAEL TODD • STEVENS, Travis Michael
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11504030-B2
Publication Date
2022-11-22
Expiration Date
Abstract
A method and system for heterogeneous event detection. Sensor data is obtained and divided into discrete data windows. Each data window is defined by and corresponds to a time period of the sensor data. A time-frequency representation over the time period is calculated for each data window. A filter mask is calculated based on the data window corresponding to the time-frequency representation. The filter mask is applied for reverting the time-frequency representation to a time representation, resulting in filtered data. Features, such as extrema or other inflection points, are identified in the filtered data. The features define events, and transforming the time-frequency representation back into the time domain emphasizes differences between more and less prominent frequencies, facilitating identification of heterogeneous events. The method and system may be applied to body movements of people or animals, automaton movement, audio signals, light intensity, or any suitable time-dependent variable.
Core Innovation
The invention provides a method for detecting heterogeneous events by receiving pressure sensor data and defining a data window over a time period of the pressure sensor data. A time-frequency representation is calculated for the time period corresponding to the data window, so that time-varying spectral content is represented for the selected window.
A filter mask is calculated based on the time-frequency representation. The filter mask comprises non-binary values proportional to a positive exponent of a magnitude of a time-frequency transform, where more prominent frequencies are emphasized and less prominent frequencies are de-emphasized.
The time-frequency representation is filtered with the filter mask to provide filtered data. From the filtered data, features are identified and a body movement event is identified with reference to the features, and the method outputs the body movement event.
Claims Coverage
The only independent claim explicitly provided is clm-00001. It includes three inventive features covering windowed time-frequency processing, non-binary adaptive filter masking based on a positive exponent of transform magnitude, and feature-based body movement event identification and output.
Windowed time-frequency event detection from pressure sensor data
Receiving pressure sensor data; defining a data window over a time period; calculating a time-frequency representation corresponding to the time period.
Non-binary adaptive filter mask from positive exponent magnitude
Calculating a filter mask based on the time-frequency representation, wherein the filter mask comprises non-binary values proportional to a positive exponent of a magnitude of a time-frequency transform, emphasizing more prominent frequencies and de-emphasizing less prominent frequencies.
Time-frequency filtering followed by feature-based body movement event identification
Filtering the time-frequency representation with the filter mask to provide filtered data; identifying features in the filtered data; identifying a body movement event with reference to the features; outputting the body movement event.
Across the provided independent claim, the inventive coverage centers on windowed time-frequency representation of pressure sensor data, an adaptive non-binary filter mask whose values are proportional to a positive exponent of a time-frequency transform magnitude, and feature extraction from filtered data to identify and output a body movement event.
Stated Advantages
Avoids arbitrary single cut-off thresholds of global low-pass/high-pass Fourier filtering.
Improves event separation under irregular/variable heterogeneous activities (e.g., gait steps).
Documented Applications
Detecting heterogeneous body movement events (including gait steps and body movement events such as walking/running/jumping) from pressure sensor data.
Wearable sensor deployments and processing contexts including shoe-inserts and other wearable sensor arrangements (e.g., wearable sensors associated with processing and event output).
Interested in licensing this patent?