Touchscreen Text Input via Trajectory Feature Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sliding input methods on touch screens, especially on small devices, suffer from low accuracy and high error rates due to the small key size and the reliance on coordinate-based detection, which increases computational complexity and power consumption, making it difficult for users to input text efficiently and accurately.
Innovation Solution
A method and system for text input by continuous sliding operation that records movement trajectories, extracts feature information such as angle and trajectory length, and filters words based on these features to provide a more accurate and efficient input process, using a combination of rough and fine filtering to reduce errors and improve user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If coordinate-based detection is used for sliding input, then the input speed is improved, but the accuracy deteriorates due to small key size and high error rate
Solution Approach 1:
The patent changes the detection parameters from simple coordinate-based detection to multi-parameter detection including trajectory length, angle, and curvature. This allows the system to distinguish between intentional key presses and sliding motions more accurately, improving input accuracy while maintaining speed.
Solution Approach 2:
The patent adds temporal and spatial dimension analysis to the detection process by examining trajectory length, angle changes, and curvature over time. This multi-dimensional approach enables more accurate differentiation between typing and sliding gestures, resolving the accuracy-speed tradeoff.
2Productivity
If the area of each key is reduced to fit more keys on small screens, then the screen efficiency is improved, but the reliability deteriorates due to increased misoperation
Solution Approach 1:
The patent implements dynamic detection that adapts to the user's input intent by analyzing trajectory characteristics in real-time. The system dynamically adjusts its recognition criteria based on the detected motion patterns, allowing small keys to be reliably distinguished despite their reduced size.
Solution Approach 2:
The system provides feedback by analyzing the complete trajectory and using curvature and angle information to confirm intended key presses. This feedback mechanism reduces misoperations by verifying that the sliding motion truly intended to press the detected key.
3Measurement precision
If complex trajectory analysis is performed to improve accuracy, then the measurement precision is improved, but the device complexity increases
Solution Approach 1:
The patent segments the trajectory analysis into distinct computational stages: basic coordinate tracking, trajectory parameter extraction (length, angle, curvature), and final recognition decision. This segmentation allows complex analysis to be performed in manageable steps, reducing overall system complexity.
Solution Approach 2:
The system performs partial trajectory analysis by focusing on key discriminative features (curvature, angle, length) rather than analyzing every aspect of the motion. This selective approach achieves high accuracy without requiring complete and exhaustive trajectory processing.
Data Source
AI summary
A method and system for text input by a continuous sliding operation is provided. Said method comprises: detecting a movement trajectory of a sensing apparatus, recording input information, and obtaining feature information from said input information, wherein said feature information includes at least one of angle information and movement trajectory length; searching a feature lexicon according to said feature information, and selecting words that satisfying condition that the angle information is matched or the movement trajectory length is matched, as matched words so as to obtain a rough screened word set; calculating a distance between the actual movement trajectory of each word and its corresponding first standard trajectory, so as to obtain a rough filtered word set; calculating a similarity between the actual movement trajectory of each word in the rough filtered word set and its corresponding second standard trajectory, and sequencing said words according to the similarity, and then outputting said words.