Touch Contact Image Classification for Palm Rejection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing touch sensor devices struggle to accurately distinguish between intentional and unintentional user contacts, particularly due to palm contacts and thumb interactions, which degrade user experience.
Innovation Solution
A neural network-based classification system that processes touch data from sensor electrodes to generate touch images, identify contact segments, and classify contacts as intentional or unintentional using neural networks, enhancing accuracy and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the touch region size is increased to provide larger interaction area, then the active usable area is improved, but palm contacts and unintentional contacts increase
Solution Approach 1:
The touch sensor data is segmented into multiple contact images, where each contact image represents a distinct contact region. The processing system divides the touch data into separate segments corresponding to different contacts (e.g., thumb contact vs. palm contact), allowing independent classification of each segment. This segmentation enables the system to distinguish intentional thumb interactions from unintentional palm contacts even when both are present on the enlarged touch surface.
2Measurement precision
If traditional touch classification methods are used, then device complexity is low, but classification accuracy between intentional and unintentional contacts deteriorates
Solution Approach 1:
A neural network classifier is introduced as an intermediary component between the touch sensor and the processing system. This neural network acts as a mediator that receives contact images and automatically classifies them as intentional or unintentional contacts based on learned patterns from training data. The neural network intermediary handles the complex classification task, improving accuracy while allowing the main processing system to focus on other functions.
Solution Approach 2:
The processing system performs preliminary actions by generating contact images and extracting relevant features before classification. The system pre-processes the touch data by creating segmented contact images, calculating contact properties (area, depth, position), and preparing the data in an optimized format for the neural network classifier. This preliminary preparation improves the efficiency and accuracy of the subsequent classification step.
3Measurement precision
If neural network based classification is implemented, then contact classification accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The system performs preliminary classification by first analyzing contact properties such as contact area, depth, and position to identify likely intentional contacts before applying the full neural network classification. This two-stage approach allows the system to quickly handle obvious cases without requiring computationally intensive neural network processing, thereby reducing overall processing time while maintaining high accuracy.
Solution Approach 2:
The touch data is segmented into multiple contact images that can be processed independently and in parallel. By dividing the classification task into separate contact segments rather than processing all touch data as a single unit, the system reduces the computational burden on the neural network and enables faster processing through parallel computation of multiple contact images.
Data Source
AI summary
An input device for classification of an input object is provided. The input device comprises a touch sensor comprising a plurality of sensor electrodes configured to obtain touch data; and a processing system. The processing system is configured to receive touch data from resulting signals from the plurality of sensor electrodes; generate a touch image based on the touch data; generate one or more contact images based on the touch image, each contact image comprising one or more first pixels from the touch image and one or more second pixels with predefined values; classify, using a neural network, a respective contact in each of the one or more contact images and generate corresponding classification results; and identify, based on the classification results, one or more classified contacts in the touch image.