Shape-Based Biometrics for Distorted-Image Subject Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image-based identification systems struggle to accurately recognize individuals under distortive conditions, such as long-range imaging, atmospheric turbulence, and clothing variations, which affect the clarity and visibility of biometric features, leading to inefficiencies and high computational costs.
Innovation Solution
The use of shape-based biometric methods, including silhouette and inverse silhouette images, combined with multi-scale representations and machine learning models like HR-Net, to generate distortion-invariant body biometrics (DIRB) and outfit regularizing biometrics (ORB), allowing for robust identification using a single or small number of images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image-based identification methods are used under distortive conditions, then recognition accuracy deteriorates, but switching to shape-based methods increases computational complexity
Solution Approach 1:
The patent extracts only the essential shape information from images by generating silhouettes and distance transformed images, removing distracting appearance details like clothing and colors. This extraction process creates simplified representations (DIRB and ORB) that retain identification-critical features while eliminating computational overhead from processing full-color images with variable appearances.
Solution Approach 2:
The patent transforms images through parameter changes including converting to grayscale, generating binary silhouettes, computing distance transforms, and creating multi-scale representations. These parameter transformations convert complex color images into simplified shape-based representations that are more robust to appearance variations and require less computational power for comparison.
2Reliability
If multiple images are used to improve recognition accuracy under distortive conditions, then identification reliability improves, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing images into standardized shape-based representations (silhouettes, distance transformed images, multi-scale features) before comparison. This preliminary transformation creates a consistent format that enables faster matching and reduces the computational burden during the actual identification process, allowing reliable recognition even with limited images.
Solution Approach 2:
The patent segments the image processing into distinct stages: silhouette generation, distance transformation, multi-scale feature extraction, and comparison. This segmentation allows each stage to be optimized independently and enables parallel processing of multiple images, reducing overall processing time while maintaining identification reliability.
3Speed
If appearance-based identification is used, then recognition speed is fast under clear conditions, but accuracy deteriorates under distortive conditions like long-range imaging and atmospheric turbulence
Solution Approach 1:
The patent converts the harmful effect of appearance variations (caused by long-range imaging, atmospheric turbulence, and clothing changes) into a benefit by deliberately removing appearance information and retaining only shape-based features. This conversion transforms distorted color images into robust silhouettes that are immune to appearance degradation, maintaining both speed and accuracy under distortive conditions.
Data Source
AI summary
Methods and systems for determining an identity of a subject based on a single-frame binary shape-capturing image extracted from distorted image of the subject and using a shape-based biometric image derived from the shape-capturing image. The shape-based biometric image includes a biometric feature of the subject and is generated by transforming the shape-capturing image to a distance transformed image and deriving a multi-scale representation of the distance transformed image. The identity of the subject can be further determined using an outfit regularizing biometric image derived from the distorted image using the shape-based biometric image. The outfit regularizing biometric includes biometric feature of the subject independent of an outfit of the subject and is generated by replacing a region of subject's boy covered by an outfit with corresponding region of the shape-based biometric image.