Living Skin Tissue Detection via Multi-Scale Pulse Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for identifying living skin tissue in video sequences face challenges such as misclassification due to non-unique human appearance features and failure with unpredictable samples, particularly when skin areas are smaller or larger than anticipated, leading to incorrect detection of objects with similar appearance.
Innovation Solution
A method that divides image frames into multiple segments and sub-sequences, analyzes pulse signals across different scales and spatial positions, and uses similarity matrices with matrix decomposition to identify areas of living skin tissue, leveraging physiological features like pulsatility for accurate differentiation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If supervised training methods using human appearance features are used to detect subjects, then detection can be performed for known samples, but the method fails with unpredictable samples and misclassifies objects with similar appearance
Solution Approach 1:
The patent changes the detection parameter from static appearance features to dynamic physiological features (pulse signals). By analyzing temporal variations in skin reflectance across multiple video frames, the system detects pulsatility characteristics that are unique to living tissue, thereby improving reliability for unpredictable samples while maintaining adaptability to different skin tones and conditions
Solution Approach 2:
The patent substitutes the mechanical/visual inspection approach with a physiological signal-based approach. Instead of relying on camera-based appearance detection, the system uses photoplethysmography (PPG) techniques to extract pulse signals from video data, replacing appearance-based classification with physiological signal analysis that is inherently more reliable for identifying living tissue
2Device complexity
If a fixed segmentation approach is used to divide image frames, then processing is simple, but the method fails when skin areas are significantly smaller or larger than anticipated
Solution Approach 1:
The patent implements dynamic segmentation where the number and size of frame segments are adaptively determined based on the detected pulse signals. The system initially divides frames into multiple segments, analyzes each for pulsatility, and then merges or refines segments based on the presence and characteristics of pulse signals, allowing the segmentation to adapt to varying skin sizes while maintaining processing efficiency
Solution Approach 2:
The patent applies multi-scale segmentation by dividing image frames into multiple frame segments at different scales. This allows the system to detect pulse signals at various resolution levels, ensuring that skin areas of different sizes can be properly identified. The segmented approach enables independent analysis of different regions while maintaining overall detection accuracy across varying skin sizes
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
This approach enables robust and scale-invariant detection of living skin tissue, improving accuracy across various skin tones, postures, and environments, while avoiding false positives from non-human objects with similar appearances.
Implementation Method 1
a video sequence of image frames captured using an imaging unit (e.g. a camera)
Implementation Method 2
techniques for performing remote photoplethysmography (remote PPG or rPPG) have been developed. These techniques enable a PPG signal to be obtained from a video sequence
Data Source
AI summary
According to an aspect, there is provided an apparatus for identifying living skin tissue in a video sequence, the apparatus comprising a processing unit configured to obtain a video sequence, the video sequence comprising a plurality of image frames; divide each of the image frames into a first plurality of frame segments; form a first plurality of video sub-sequences, each video sub-sequence comprising a frame segment from the first plurality of frame segments for two or more of the image frames; divide each of the image frames into a second plurality of frame segments, the second plurality of frame segments comprising a greater number of frame segments than the first plurality of frame segments; form a second plurality of video sub-sequences, each video sub-sequence comprising a frame segment from the second plurality of frame segments for two or more of the image frames; analyze the video sub-sequences in the first plurality of video sub-sequences and the second plurality of video sub-sequences to determine a pulse signal for each video sub-sequence; and analyze the pulse signals to identify areas of living skin tissue in the video sequence.