RPPVSM Pattern Matching for Obscured Face Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies face challenges in accurately identifying individuals from digital media, particularly in cases of obscured faces and lack of traditional biometric data, such as in child pornography cases, where relatively permanent pigmented or vascular skin mark patterns (RPPVSM) have not been effectively utilized for identification.
Innovation Solution
The method involves identifying and matching RPPVSM patterns in images using a point matching model, calculating correspondence probabilities, and employing a theoretical and experimental approach to determine the likelihood of identification, with the use of nevus patterns as an example of RPPVSM, to positively identify individuals from digital evidence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional biometric data (fingerprints, palm prints) is used for identification, then identification accuracy is improved, but applicability to obscured faces and digital media is worsened
Solution Approach 1:
The patent extracts and utilizes RPPVSM patterns from digital media images as a new biometric identifier. By taking out the skin mark patterns from the overall image data and processing them independently, the system achieves identification capability that works even when facial features are obscured, thus resolving the contradiction between identification accuracy and adaptability to obscured faces.
Solution Approach 2:
The patent introduces RPPVSM patterns as an intermediary biometric feature between the digital media evidence and the identification process. These skin mark patterns serve as a mediator that can be extracted from various digital media types (images, video frames) and processed to provide reliable identification, bridging the gap between traditional biometrics and modern digital evidence.
2Adaptability or versatility
If RPPVSM patterns are used for identification, then adaptability to digital media is improved, but measurement precision and identification reliability are worsened
Solution Approach 1:
The patent replaces manual visual inspection and traditional fingerprint analysis with an automated computational system. The system uses image processing algorithms, pattern recognition, and statistical models to automatically detect, extract, and match RPPVSM patterns, thereby improving both adaptability to digital media and identification reliability through consistent, reproducible processing.
Solution Approach 2:
The patent transforms the identification process by changing the parameters being measured from traditional biometric features to RPPVSM pattern characteristics (location, morphology, distribution). By changing these parameters and using them in a standardized computational framework, the system achieves both high adaptability to various digital media formats and reliable identification through statistical validation.
3Device complexity
If point matching model is used to match RPPVSM locations, then identification process is simplified, but measurement precision is worsened
Solution Approach 1:
The patent segments the identification process into distinct stages: detection of RPPVSM patterns, extraction of location and morphological data, alignment of patterns between images, and statistical evaluation of matches. By segmenting the complex task into manageable steps, the system maintains measurement precision at each stage while keeping the overall process organized and manageable.
Solution Approach 2:
The patent incorporates feedback mechanisms through statistical evaluation and probability calculations. After matching RPPVSM points between images, the system calculates correspondence probabilities and uses this feedback to validate or reject potential matches. This feedback loop ensures measurement precision is maintained while simplifying the decision-making process for identification.
Data Source
AI summary
Embodiments include methods, devices, software, and systems for identifying a person based on relatively permanent pigmented or vascular skin mark (RPPVSM) patterns in images. Locations of RPPVSMs in different images of people are point matched, and a correspondence probability that the point matched RPPVSMs are from different people is calculated. Other embodiments are also described. Other embodiments are also described and claimed.