Multi-Touch Detection Using Local Maxima Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multi-touch detection methods are limited by the number of touches they can detect, are not robust in the presence of noise, and require high processing power, making them unsuitable for applications like hand/stylus writing/painting and use by multiple people.
Innovation Solution
A method that uses a capacitive sensing device with a non-transitory computer readable medium to detect and track an arbitrary number of touches, employing image processing techniques such as local maxima search, binary morphological operations, and linear assignment problems to achieve robust and flexible gesture recognition with low processing requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional multi-touch detection methods are used, then the system can detect touches, but the number of detectable touches is limited and processing power requirements increase
Solution Approach 1:
The patent segments the touch detection process into distinct phases: calibration phase (storing reference values) and detection phase (comparing current readings to references). This segmentation allows the system to handle arbitrary numbers of touches by processing them in manageable steps rather than requiring complex simultaneous analysis of all touches.
Solution Approach 2:
The patent performs preliminary calibration actions before actual touch detection, storing reference capacitance values for each sensor element in the absence of touches. This preliminary action enables subsequent detection to focus only on detecting deviations from known baselines, reducing processing complexity regardless of the number of touches.
2Measurement precision
If complex image processing techniques are used to improve touch detection accuracy, then measurement precision improves, but processing power requirements increase
Solution Approach 1:
The patent extracts and processes only the most relevant features for touch detection - specifically comparing current capacitance readings against calibrated reference values to identify touches. This selective extraction avoids the computational burden of processing entire image sequences while maintaining precision in determining touch positions.
Solution Approach 2:
The patent applies local quality by focusing processing resources on identifying and analyzing touch regions rather than uniformly processing the entire sensor array. By concentrating analysis on areas where capacitance changes indicate touches, the system achieves high measurement precision with reduced overall processing requirements.
3Reliability
If robust noise filtering is applied to handle noisy input data, then reliability of detection improves, but processing time increases
Solution Approach 1:
The patent cushions against noise by establishing calibrated reference values during a preliminary phase when no touches are present. These pre-established references serve as a buffer against noisy variations during actual touch detection, allowing the system to distinguish between normal noise and genuine touch signals without requiring complex real-time filtering.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables true multi-touch detection without limitations on the number of touches, suitable for hand/stylus writing/painting applications, and allows flexible gesture recognition with reduced processing power, handling noisy input data effectively.
Implementation Method 1
capacitive sensing device comprising: an array comprising a first number of rows and a second number of columns of sensing capacitors
Implementation Method 2
If an object with good dielectric properties, e.g. a finger, is approaching the device, the capacitance of the underlying capacitors is changed
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
A method for detecting an arbitrary number of touches from an input image of a multi-touch device comprising the following steps: -processing said input image and obtaining a processed image; -segmentation by thresholding of said processed image (8000) and obtaining a segmented image; -identifying regions of the segmented image; -finding local maxima, each local maximum being of size one pixel on a sub-region inside each region; -determining at least one touch position based on said local maxima.