Optical Flow Breath-Rate Sensing with Low-Memory Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing breath rate detection systems using cameras face high processing and memory requirements, and may infringe on privacy due to image capture, necessitating a more efficient and privacy-protecting method.
Innovation Solution
Utilizing an optical flow sensor to capture image frames, determine movement values, and select a subset of frames with the highest movement values to calculate breath rate, employing AI algorithms and neural networks for human detection and region of interest identification, while minimizing memory footprint.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If camera-based breath rate detection systems are used, then breath rate can be detected, but processing requirements and memory requirements increase significantly
Solution Approach 1:
The patent extracts only the essential movement information from full image frames by using optical flow sensors to capture motion data specifically. Instead of processing complete images, the system extracts movement values and selects only frames with significant movement, thereby reducing processing requirements while maintaining breath rate detection capability
Solution Approach 2:
The patent uses optical flow sensors to create simplified motion representations (movement values) that copy only the essential breathing-related information from the visual scene. This copying approach replaces full image processing with lightweight motion metric calculation, reducing computational complexity while preserving detection accuracy
2Measurement precision
If camera-based breath rate detection systems are used, then breath rate can be detected, but memory requirements increase significantly
Solution Approach 1:
The system extracts only movement values from image frames using optical flow sensing, storing only these scalar metrics rather than entire image frames. This extraction reduces memory footprint from megabytes per frame to bytes per movement value, while maintaining the ability to detect breath rate through movement analysis
Solution Approach 2:
The patent processes and discards individual image frames sequentially, keeping only the current frame in memory at any time. By using optical flow to compute movement values on-the-fly and immediately discarding processed frames, the system achieves breath rate detection with minimal memory footprint, treating each frame as a temporary, disposable unit
3Loss of information
If full image frames are processed for breath rate detection, then comprehensive motion information is available, but processing time and computational load increase
Solution Approach 1:
The patent extracts only the movement value from each image frame using optical flow sensors, discarding all other visual information. This selective extraction maintains the essential motion information needed for breath rate detection while eliminating redundant data, thereby increasing processing speed without significant loss of relevant motion information
Solution Approach 2:
The system performs partial processing by calculating only movement values rather than analyzing complete image content. This partial action approach processes only the necessary motion metrics for breath detection, achieving sufficient motion information analysis with dramatically reduced computational load and faster processing throughput
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
Achieves contactless breath rate recognition with reduced memory and processing demands, providing privacy protection and flexible monitoring without large memory or power consumption.
Implementation Method 1
systems, devices, and methods for determining breath rate using optical flow sensing
Data Source
AI summary
Various embodiments provide systems, devices and methods for determining breath rate by capturing a plurality of image frames from a region of interest, determining a subset of image frames that include image frames having largest movement values in the plurality of image frames, determining a first image frame in the subset of image frames and a last image frame in the subset of image frames. In various embodiments, the systems, devices and methods determine the breath rate using the first image frame and the last image frame in the subset of image frames and a rate of capturing the plurality of image frames.