Touch Panel Image Recognition for Multi-Input Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Information processing apparatuses with handwriting input functions face challenges such as time lag in displaying written content, erroneous recognition, and decreased recognition precision, especially during multi-touch inputs.
Innovation Solution
The apparatus incorporates a camera, an image recognition module, and a display controller, utilizing a first and second narrowing module to extract and select image data for recognition based on timing and positional information, reducing processing time and space, and implementing mechanisms for automatic recovery from erroneous recognition and maintaining precision during simultaneous inputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image recognition is performed to recognize the color of a writing tool, then the displayed content can be attached with the correct color information, but a time lag occurs until the written content is displayed after writing
Solution Approach 1:
The system performs preliminary image recognition processing in the background before the user completes their input. The image recognition module continuously monitors the touch panel surface and pre-processes images to identify writing tool colors, so that when the user releases the touch panel, the color information is already prepared and can be immediately applied to the displayed content without causing display lag.
Solution Approach 2:
The image recognition process runs continuously in the background during the entire writing process, rather than being triggered only after writing completes. This continuous processing ensures that color recognition is performed without interruption, maintaining both accuracy and responsiveness by constantly updating the recognized color information as the user writes.
2Device complexity
If conventional image recognition is used without careful consideration of erroneous recognition, then the system is simpler to implement, but erroneous recognition occurs requiring users to repeatedly erase and rewrite content
Solution Approach 1:
The system implements a feedback mechanism where the recognized color information is continuously monitored and verified. When erroneous recognition is detected (such as when the recognized color does not match the actual writing tool color or when recognition results are inconsistent across multiple samples), the system automatically triggers correction procedures including re-recognition and error recovery mechanisms, thereby maintaining high reliability without significantly increasing system complexity.
Solution Approach 2:
The system dynamically adjusts recognition parameters such as color threshold values, recognition sensitivity, and processing algorithms based on the current writing context and environmental conditions. By adapting these parameters in real-time, the system optimizes recognition accuracy for different scenarios while maintaining a relatively simple overall system architecture.
3Device complexity
If a single recognition process is used for simultaneous inputs, then the system is simpler to manage, but recognition precision is markedly degraded when multiple users write in close positions
Solution Approach 1:
The system segments the touch panel surface into multiple recognition zones and assigns separate recognition processes to each zone when multiple inputs are detected. This spatial segmentation allows each writing tool to be independently recognized within its own zone, preventing cross-contamination of recognition data and maintaining high recognition precision even when multiple users write in close proximity on the same panel.
Solution Approach 2:
The system introduces a temporal dimension to the recognition process by processing multiple inputs in a time-sequenced manner while maintaining spatial separation. When simultaneous inputs are detected, the system assigns unique identification markers and processes each input stream separately through dedicated recognition channels, effectively adding a dimensional layer to distinguish between multiple concurrent writing actions.
Data Source
AI summary
According to one embodiment, an information processing apparatus includes a display with a touch panel, a camera and an image recognition module. The image recognition module recognizes a type of an object used for inputting to the touch panel in an image acquired by the camera. The image recognition module comprises a first narrowing module and a second narrowing module. The first narrowing module extracts an image to be used for recognizing the type of the object from acquired images based on timing information indicating timing of the inputting. The second narrowing module selects a recognition range to be used for recognizing the type of the object from a photographing region of the extracted image based on positional information indicating an input position on the touch panel. The timing information and the positional information is included in the input information obtained from the touch panel.