Knuckle Touch Recognition Using ACC-CAP Feature Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional knuckle touch detection algorithms suffer from low identification rates and high false touch rates, failing to effectively distinguish knuckle operations from other touch types on touch panels.
Innovation Solution
A method involving the extraction of ACC, confidence score, and touch features from respective signals, followed by feature fusion and classification using neural network models to improve knuckle operation identification accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional knuckle touch detection algorithms are used, then the system can identify knuckle operations, but the identification rate is low and false touch rate is high
Solution Approach 1:
The patent segments the touch detection process into multiple independent feature extraction stages: ACC signal processing, CAP signal processing, and touch signal processing. Each stage extracts specific features (acceleration features, capacitance features, touch pressure features) that are then combined for comprehensive knuckle operation identification, improving both accuracy and reliability
Solution Approach 2:
The patent merges multiple signal sources (ACC, CAP, and touch signals) and their respective features into a unified feature set for knuckle operation identification. This combination of multiple feature types enables more accurate distinction between knuckle operations and other touch types, reducing false touch rates while improving identification accuracy
2Adaptability or versatility
If conventional single-signal detection methods are used, then the algorithm complexity is low, but the generalization performance is insufficient
Solution Approach 1:
The patent transitions from single-signal detection to multi-signal multi-feature detection by adding temporal dimensions (acceleration over time), capacitive dimensions (capacitance values), and pressure dimensions (touch pressure). This multi-dimensional approach significantly improves algorithm generalization performance across different knuckle operation scenarios while maintaining manageable complexity through systematic feature extraction
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Embodiments of this application relate to the field of artificial intelligence technologies, and provide a knuckle operation identification method and an electronic device. According to solutions of this application, when a knuckle performs an operation on different touch regions of a touch panel, an ACC signal may have different change trends, and therefore it is proposed to extract a touch feature from a touch signal for determining a contact area and a contact location, and to extract an ACC feature from the ACC signal for determining a magnitude of a touch on a screen. In addition, a CAP binary classification model is used at the front, to extract a confidence score indicating a correlation to a knuckle from a CAP signal. Then, feature fusion is performed on the ACC feature, the confidence score, and the touch feature, and a fused feature is input into a knuckle classification model, to obtain a better classification effect.