Multi-Touch Gesture Recognition via Temporal Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multi-touch gesture recognition systems cannot perform continuous recognition of multi-touch gestures, as they require users to lift their fingers off the device between gestures and cannot interpret dynamic changes in contact points, limiting the ability to perform a sequence of temporally contiguous gestures.
Innovation Solution
A method and system for continuous recognition of multi-touch gestures that allows users to interact without lifting their fingers, using spatial and temporal segmentation of input data, and employing a multi-agent approach with finite state machines to distinguish between gestures, enabling the recognition of dynamic multi-touch gestures and sequences of gestures performed concurrently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional multi-touch gesture recognition systems are used, then gesture recognition is possible, but continuous recognition of sequences of gestures without lifting fingers is not achievable
Solution Approach 1:
The patent segments the continuous input data stream into discrete gesture events by identifying start and end points based on contact point changes. This segmentation allows the system to process complex continuous interactions as a sequence of manageable gesture units, enabling continuous recognition without overwhelming system complexity.
Solution Approach 2:
The system dynamically adapts its recognition state based on real-time contact point changes. By transitioning between different recognition states (idle, gesture in progress, gesture completed) based on dynamic input conditions, the system enables continuous gesture sequences while maintaining manageable complexity through state-based control.
2Measurement precision
If multi-touch input devices are used, then multiple contact points can be detected, but interpretation of dynamic changes in contact points for continuous gestures is limited
Solution Approach 1:
The system continuously monitors contact point changes and uses this feedback to dynamically adjust gesture recognition. By comparing current contact point states with previous states, the system can identify gesture boundaries and interpret dynamic changes, enabling versatile gesture sequence recognition while maintaining precise contact point detection.
Solution Approach 2:
The patent establishes predetermined criteria for gesture start and end conditions based on contact point changes. By pre-defining these transition conditions, the system can quickly and accurately interpret dynamic contact point changes as meaningful gesture events, enhancing both measurement precision and adaptability.
3Reliability
If users must lift fingers between gestures, then gesture boundaries are clear, but interaction fluidity and naturalness are reduced
Solution Approach 1:
The system dynamically determines gesture boundaries based on contact point changes rather than requiring physical finger lifts. By adapting boundary detection to the actual input dynamics, the system maintains reliable gesture differentiation while enabling fluid, natural continuous interaction sequences.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
The present invention extends to methods, systems, and computer program products for processing multi-touch inputs. It provides a method for continuous recognition of multi-touch gestures performed by one or more users on at least two multi-touch input devices. The invention in one aspect allows users to start interacting with one multi-gesture and subsequently perform any other multi-touch gestures, if they want it, without needing to lift all their fingers off the multi-touch input devices between each multi-touch gesture.