Heartbeat Rate Detection from Video Using Adaptive Facial Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current remote photoplethysmography technologies struggle to accurately and reliably determine heartbeat rates from video recordings of naturally interacting human subjects in dynamic settings, due to limitations such as subject mobility, varying lighting conditions, and the need for specialized equipment.
Innovation Solution
The development of computationally efficient methods and systems that analyze video recordings to determine heartbeat rates by employing face recognition to identify adaptive regions of interest, extracting light intensity signals, and processing them using spectral and peak-finding methods to produce accurate heartbeat rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If remote photoplethysmography analyzes video recordings to determine heartbeat rate, then non-invasive monitoring is achieved, but measurement precision deteriorates in dynamic settings with subject mobility and varying lighting
Solution Approach 1:
The patent segments the video signal processing into distinct stages: raw video acquisition, preprocessing (motion compensation, illumination normalization), photoplethysmographic signal extraction, and heartbeat rate determination. This segmentation allows each stage to be optimized independently, maintaining measurement precision while preserving non-invasive monitoring capabilities.
Solution Approach 2:
The patent introduces intermediate processing steps including motion compensation algorithms and illumination normalization techniques that act as mediators between the raw video signal and the final heartbeat rate measurement. These intermediaries correct for environmental disturbances before the actual physiological signal extraction, thereby maintaining accuracy in dynamic settings.
2Measurement precision
If specialized equipment is used for accurate heartbeat rate monitoring, then measurement precision improves, but device complexity and cost increase
Solution Approach 1:
The patent uses standard video recording equipment to capture physiological information, effectively creating an optical copy of the photoplethysmographic signal that would normally require specialized sensors. By processing this video copy through specialized algorithms rather than requiring specialized hardware, the system achieves accurate heartbeat rate monitoring without complex equipment.
Solution Approach 2:
The patent replaces the mechanical/optical system of specialized photoplethysmography sensors with a computational system that processes standard video footage. Instead of using dedicated hardware to detect blood volume changes, the system uses image processing algorithms to extract the same physiological information from conventional video cameras, thereby reducing device complexity.
3Measurement precision
If subjects are kept immobile and in static environments for accurate remote photoplethysmography, then measurement precision improves, but adaptability to natural settings deteriorates
Solution Approach 1:
The patent implements dynamic motion compensation that adapts to subject movement in real-time. Rather than requiring static conditions, the system continuously adjusts for changing camera-subject geometry, varying illumination angles, and subject position changes. This dynamic adaptation allows accurate heartbeat rate determination in natural, mobile settings.
Solution Approach 2:
The patent changes the processing parameters adaptively based on environmental conditions. Illumination normalization parameters, motion compensation vectors, and signal extraction windows are dynamically adjusted according to the actual lighting conditions and subject movement patterns observed in each recording session, enabling accurate measurements across diverse natural settings.
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
These methods provide robust and accurate heartbeat rate determination for human subjects in various natural settings, using non-specialized video recording equipment, and are capable of handling subjects in motion and under varying lighting conditions.
Implementation Method 1
subtle changes in skin coloration that oscillate at frequencies close to the heartbeat-rate frequency
Implementation Method 2
remote photoplethysmography based on analysis of video recordings generally relies on subtle changes in skin coloration
Data Source
AI summary
The current document is directed to computationally efficient, robust, and accurate methods and systems that determine the heartbeat rate of human subjects, patients, and/or participants monitored by a video-recording-capable device or system by analyzing video recordings of human subjects, patients, and/or participants, either in real time or following video recording. The currently disclosed methods and systems employ face-recognition methods to identify adaptive regions of interest corresponding to facial-skin areas. Spatial averages of the intensities of light components with different frequencies reflected from the identified adaptive regions of interest are extracted from the video recording as intensity signals. The intensity signals are then analyzed to extract detrended, band-pass-filtered oscillating light-intensity signals that are processed by spectral and peak-finding methods that each produces a determined heartbeat rate for each of multiple time intervals.