Video Pulse Waveform Detection Using 3D CNN ROI Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video-based techniques for pulse waveform detection are inaccurate, require physical contact, and impose constraints on subject movement and posture, making them unsuitable for real-world applications.
Innovation Solution
A system utilizing a 3-dimensional convolutional neural network (3DCNN) processes spatial and temporal dimensions of video streams to accurately detect pulse waveforms without contact, allowing for minimal subject movement and posture constraints, by spatially locating regions of interest and applying dilations to 3D kernels based on depth.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If video based techniques are used for pulse waveform detection, then physical contact is eliminated, but measurement accuracy deteriorates
Solution Approach 1:
The patent transitions from traditional 2D image processing to 3D volumetric processing by incorporating depth information from structured light or time-of-flight sensors. This third dimension enables more robust pulse detection by analyzing blood volume changes in three-dimensional space, significantly improving measurement accuracy while maintaining contactless operation
Solution Approach 2:
The system merges multiple sensing modalities including visible light imaging, infrared thermography, and depth sensing into a unified pulse detection framework. This multi-modal approach compensates for the weaknesses of individual techniques and enhances overall measurement precision through data fusion
2Ease of operation
If existing video based methods are used, then contactless measurement is achieved, but reliability deteriorates due to subject movement and posture constraints
Solution Approach 1:
The system implements dynamic tracking algorithms that continuously adapt to subject movement and changing posture. The 3D spatial information enables real-time identification and tracking of stable vascular regions, maintaining measurement reliability even when the subject moves or changes position during the measurement process
Solution Approach 2:
The patent dynamically adjusts processing parameters such as region of interest selection, filtering characteristics, and analysis window size based on detected movement levels and signal quality. This adaptive parameter adjustment maintains measurement reliability across varying conditions without requiring subject stillness
3Measurement precision
If 3D convolutional neural network is used for spatial and temporal processing, then pulse waveform detection accuracy is improved, but computational complexity increases
Solution Approach 1:
The 3D video data is segmented into multiple temporal frames and spatial regions, with the convolutional neural network processing these divided segments independently before aggregation. This segmentation reduces the computational burden on any single processing operation while maintaining overall detection accuracy through comprehensive analysis of all segments
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
The system provides accurate pulse waveform detection with minimal subject constraints, enabling reliable heart rate estimation and variability analysis over short video durations, improving model performance without increasing computational requirements.
Implementation Method 1
The change in reflected light from the skin's surface, because of light absorption of blood, is very minor compared to those caused by changes in illumination
Data Source
AI summary
The video based detection of pulse waveform includes systems, devices, methods, and computer-readable instructions for capturing a video stream including a sequence of frames, processing each frame of the video stream to spatially locate a region of interest, cropping each frame of the video stream to encapsulate the region of interest, processing the sequence of frames, by a 3-dimensional convolutional neural network, to determine the spatial and temporal dimensions of each frame of the sequence of frames and to produce a pulse waveform point for each frame of the sequence of frames, and generating a time series of pulse waveform points to generate the pulse waveform of the subject for the sequence of frames.