Lip Shape Variation Handling in Biometric Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing living body identification methods using lip language recognition face inaccuracies due to variations in lip shapes when uttering the same character in different languages or dialects, leading to incorrect determination of a living body.
Innovation Solution
Providing lip language prompt information with at least two target characters of the same or opposite lip shapes, or with lip shape similarity within a preset range, and detecting lip changes in collected frame pictures to accurately determine if the object is a living body.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single character is used for lip language prompt information, then the identification process is simple, but the identification accuracy deteriorates due to lip shape variations across different languages and dialects
Solution Approach 1:
The patent segments the lip language prompt information into multiple characters instead of using a single character. Each character provides a different lip shape pattern, and the system evaluates multiple characters to determine if the captured lip movements match any of them. This segmentation allows the system to handle lip shape variations across different languages and dialects while maintaining identification accuracy.
Solution Approach 2:
The patent changes the parameter of lip shape similarity by introducing a preset range threshold. Instead of requiring exact matching of lip shapes, the system determines whether the lip shape similarity falls within an acceptable range. This parameter adjustment enables the system to accommodate natural variations in lip shapes when uttering the same character in different languages or dialects.
2Reliability
If lip shape matching is made strict to ensure accuracy, then false positives are reduced, but false negatives increase due to natural lip shape variations
Solution Approach 1:
The patent introduces a preset range parameter for lip shape similarity to balance strictness and flexibility. By defining an acceptable similarity range rather than requiring exact matches, the system maintains high reliability while adapting to natural lip shape variations across different languages and dialects. This parameter adjustment prevents both false positives and false negatives.
Solution Approach 2:
The patent makes the lip language prompt information universal by including multiple characters that can be uttered in different languages and dialects. The system is designed to handle various linguistic contexts, making it adaptable to different users while maintaining consistent identification reliability through the preset similarity range.
Data Source
AI summary
The present disclosure discloses a living body identification method, an information generation method, and a terminal, and belongs to the field of biometric feature recognition. The method includes: providing lip language prompt information, the lip language prompt information including at least two target characters, and the at least two target characters being at least one of: characters of a same lip shape, characters of opposite lip shapes, or characters whose lip shape similarity is in a preset range; collecting at least two frame pictures; detecting whether lip changes of a to-be-identified object in the at least two frame pictures meet a preset condition, when the to-be-identified object reads the at least two target characters; and determining that the to-be-identified object is a living body, if the preset condition is met. The present disclosure resolves a problem in the related technology that even if a to-be-identified object performs an operation according to lip language prompt information, a terminal may incorrectly determine that the to-be-identified object is not a living body, and achieves an effect that the terminal can accurately determine whether the to-be-identified object is a living body and improve determining accuracy.