Host-Based Touch Processing Center of Mass Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional touch processing algorithms face issues with suboptimal determination, aliasing, and low processing power, leading to biased and jittery touch location estimation and inaccurate shape parameter calculation due to treating capacitive panel measurements as point masses and low pixel density.
Innovation Solution
The implementation of a parallel processing system that receives touch input data points, determines a center of mass estimate, and performs downsampling to differentiate between coarse and fine objects, using maximum likelihood estimation to improve touch point positioning and shape estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If center of mass estimation is used for touch location determination, then the processing is computationally simple, but the estimation is suboptimal and subject to aliasing causing bias and high jitter
Solution Approach 1:
The patent segments the touch processing into multiple stages: initial center of mass estimation, followed by iterative refinement using the measured capacitance values and kernel function matching. This segmentation allows the system to start with a simple computational approach and progressively improve accuracy through additional processing steps.
Solution Approach 2:
The patent performs preliminary center of mass estimation to obtain an initial touch location estimate, which then serves as the starting point for subsequent refinement iterations. This preliminary action provides a good initial guess that reduces the computational burden of the more complex refinement algorithms.
2Device complexity
If low pixel density is used in capacitive touch panel, then the device complexity is reduced, but the shape parameter estimation becomes inaccurate
Solution Approach 1:
The patent changes the processing parameters by applying kernel function matching and iterative optimization algorithms that can extract accurate shape parameters from low-density pixel data. Instead of relying on high pixel density, the system transforms the problem through mathematical modeling and parameter optimization.
Solution Approach 2:
The patent substitutes the mechanical approach of increasing pixel density with a computational approach using kernel function matching and capacitance distribution analysis. This replaces the need for finer physical sampling with sophisticated signal processing and mathematical modeling.
3Use of energy by moving object
If conventional touch processing algorithms are used, then the processing power requirement is low, but the determination is suboptimal leading to biased results
Solution Approach 1:
The patent implements dynamic processing where the level of computational effort can be adjusted based on the specific touch event characteristics. The system starts with simple center of mass calculation and only performs more computationally intensive kernel matching and iterative refinement when necessary to achieve the desired accuracy level.
Solution Approach 2:
The patent introduces kernel function matching as an intermediary step between simple center of mass estimation and final touch determination. This intermediary process uses the measured capacitance values and a modeled kernel function to bridge the gap between crude initial estimation and accurate final results, improving reliability without requiring full computational power from the start.
Data Source
AI summary
Systems and methods are disclosed for determining a touch position from data received from a touch panel. In one implementation, an apparatus for processing a touch input signal includes a processor that may be configured for parallel processing, a touch device, a memory, operably connected to the processor, and configured to store processor instructions. The processor instructions can configure the processor to receive a plurality of data points corresponding to a plurality of touch events on the touch device, determine a center of mass estimate from the plurality of data points, determine a search radius of center of mass estimate, and determine an optimal touch point position based on the center of mass estimate and the search radius using the processor.