Gesture Shaking Recognition via Tracking Point Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for suppressing hand shaking in virtual reality and gesture recognition applications are inefficient, leading to unstable calculations and poor human-computer interaction experiences due to the need for full-range calculations.
Innovation Solution
A gesture shaking recognition method that acquires two adjacent frames of gesture images, selects a tracking point, determines the positional information of the maximum value pixel point, and assesses gesture shaking based on differences between tracking and pixel point positions, using thresholds to differentiate between shaking and movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full range calculation is used to suppress hand shaking, then gesture recognition accuracy is improved, but calculation time increases significantly
Solution Approach 1:
The patent divides the gesture recognition process into distinct stages: hand shaking detection phase and gesture recognition phase. During hand shaking detection, the system calculates the standard deviation of pixel values in the hand region to identify shaking states. When shaking is detected, the system suppresses it by adjusting the gesture recognition algorithm, otherwise it performs full gesture recognition. This segmentation allows the system to use computationally intensive full-range calculation only when necessary, rather than continuously.
Solution Approach 2:
The patent dynamically changes the recognition parameters based on the detected hand shaking state. When hand shaking is detected (standard deviation exceeds threshold), the system switches to a suppression mode with adjusted parameters, and when no shaking is detected, it switches to normal gesture recognition mode. This parameter switching allows the system to optimize between accuracy and computational efficiency based on real-time conditions.
2Measurement precision
If full range calculation is used to suppress hand shaking, then gesture recognition accuracy is improved, but device complexity increases
Solution Approach 1:
The patent segments the calculation process into a simple hand shaking detection stage (calculating standard deviation of pixel values in hand region) and a gesture recognition stage. This segmentation avoids the need for continuous full-range calculations, thereby reducing overall computational complexity while maintaining accuracy when needed.
Solution Approach 2:
The system performs self-assessment of hand shaking state through automated standard deviation calculation and threshold comparison, then self-adjusts the gesture recognition parameters accordingly. This self-service mechanism eliminates the need for complex external control systems or manual intervention, simplifying the overall device architecture while maintaining high recognition accuracy.
3Loss of time
If single standard suppression is used, then calculation time is reduced, but calculation stability deteriorates
Solution Approach 1:
The patent employs dynamic parameter adjustment based on the detected hand shaking state. When shaking is detected, the system switches to suppression parameters; when not detected, it uses normal recognition parameters. This dynamic adaptation ensures calculation stability by appropriately selecting parameters for each condition, avoiding the instability that would result from using a single fixed standard.
Data Source
AI summary
A gesture shaking recognition method includes acquiring two adjacent frames of gesture images of a gesture (S100), selecting a tracking point in each of the two adjacent frames of gesture images (S200), determining positional information of a maximum value pixel point in each of the two adjacent frames of gesture images based on the tracking point (S300), and determining whether gesture shaking occurs (S400) based on the positional information of the maximum value pixel point in each of the two adjacent frames of gesture images and positional information of the tracking point in each of the two adjacent frames of gesture images. The tracking point in each of the two adjacent frames of gesture images may correspond to a same position on the gesture.