Touchpad Multipoint Positioning Using Local Maxima Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-touch technologies face challenges in accurately distinguishing multiple fingers, especially when they are close to each other, leading to control errors due to asymmetry in finger area separation and centroid calculation in capacitive and optical touch panels.
Innovation Solution
A multipoint positioning method for touchpads that uses two-dimensional data to identify local maxima and object areas, simplifying the positioning process by comparing these against thresholds to accurately determine contact positions without relying on centroid calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If valley-based separation method is used to identify multiple fingers, then finger areas can be separated, but asymmetry in separated finger areas causes positioning deviation and control error
Solution Approach 1:
Instead of using the valley (minimum point) between peaks to separate finger areas, the patent inverts the approach by using the local maxima (peak points) themselves as the separation criterion. The method identifies multiple local maxima in the touch signal and uses their positions and characteristics to determine finger locations, thereby avoiding the asymmetry problem caused by valley-based separation.
Solution Approach 2:
The patent changes the parameter used for finger separation from valley position to local maxima position. By detecting and analyzing the characteristics of local maxima (such as their amplitude, position, and temporal evolution), the system achieves more accurate and symmetric finger area separation, eliminating the positioning deviation that occurs with valley-based methods.
2Ease of operation
If centroid calculation is used to locate finger position, then finger position can be obtained, but positioning deviation occurs leading to control error
Solution Approach 1:
The patent extracts the essential information needed for positioning directly from the local maxima characteristics rather than calculating the centroid of the entire finger area. By taking out the position and amplitude information from the local maxima points themselves, the method obtains finger position data without performing complex centroid calculations, thereby maintaining simplicity while improving accuracy.
3Device complexity
If one-dimensional information is used for multi-touch detection, then processing is simplified, but multiple fingers cannot be distinguished in specific conditions
Solution Approach 1:
The patent transitions from one-dimensional touch signal analysis to two-dimensional analysis by considering both the spatial distribution and temporal characteristics of local maxima. The method analyzes the two-dimensional matrix data structure and tracks the evolution of local maxima over time, adding a temporal dimension to the detection process. This enables the system to distinguish multiple fingers even when they are close together or overlapping, while maintaining relatively simple processing through the use of local maxima identification.
Data Source
AI summary
A multipoint positioning method for a touchpad including the steps of: scanning a touchpad to retrieve two-dimensional data; calculating an object area and a number of maxima of local maxima in the two-dimensional data; comparing the object area with an area threshold when the number of maxima is larger than 1; and identifying positions of the local maxima as a plurality of contact positions when the object area is larger than or equal to the area threshold.