Bi-directional Handwriting Recognition with Combined Dictionary
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current handwriting recognizers are primarily designed for either left-to-right or right-to-left languages, making it difficult for multi-lingual users to input and recognize mixed text combinations in both formats.
Innovation Solution
A bi-directional handwriting recognition system that uses a combined dictionary, including a dictionary for one direction and its reverse, to recognize and correct mixed left-to-right and right-to-left text, with an inverse bi-directional process to convert text from visual to logical order for seamless integration with other applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a handwriting recognizer is designed for a single direction (left-to-right or right-to-left), then the recognition accuracy for that specific language is improved, but the ability to recognize mixed-language text is lost
Solution Approach 1:
The patent merges multiple directional dictionaries (left-to-right and right-to-left) into a single combined dictionary structure. This allows the handwriting recognizer to handle both directional languages simultaneously without sacrificing accuracy for either language, resolving the contradiction between single-language precision and multi-language versatility.
Solution Approach 2:
The combined dictionary enables the recognizer to serve multiple language directions with a single unified structure. The system can recognize both left-to-right and right-to-left scripts in the same input stream, making the recognizer universal for multi-lingual users while maintaining directional accuracy.
2Adaptability or versatility
If a combined dictionary is used to support multiple directions, then the versatility for mixed-language recognition is improved, but the device complexity increases
Solution Approach 1:
The combined dictionary is segmented into distinct directional sections (left-to-right and right-to-left) that can be independently managed. This segmentation allows the system to handle complexity by organizing data in manageable chunks while maintaining the ability to process mixed-language input efficiently.
3Ease of operation
If handwriting text is displayed in visual order, then the user can easily read and correct characters, but the text cannot be directly integrated with applications that require logical order
Solution Approach 1:
The system introduces an intermediary conversion process that translates text from visual order (how it appears on screen for user correction) to logical order (how applications expect to process it). This intermediary layer resolves the contradiction by allowing users to interact with text in its natural visual display order while maintaining compatibility with application requirements.
4Adaptability or versatility
If an inverse bi-directional process is implemented to convert visual to logical order, then the integration with applications is improved, but the processing time increases
Solution Approach 1:
The system performs preliminary directional analysis during the initial handwriting recognition phase, determining the text direction before final processing. This preliminary action reduces the complexity and time required for subsequent conversion to logical order, as the inverse bi-directional process only needs to handle the conversion rather than analyzing directionality from scratch.
Data Source
AI summary
Various technologies and techniques are disclosed for providing bi-handwriting directional handwriting recognition and correction. A combined handwriting recognizer is provided that supports left-to-right and right-to-left language recognition by using a combined dictionary. The combined dictionary includes a dictionary from a language in a first direction, along with a backwards version of a dictionary from a language in a second direction. The combined recognizer is used with the combined dictionary to generate a most likely recognition result for mixed direction hand written input received from a user. Character by character correction is provided for mixed left-to-right and right-to-left text. The most likely recognition result is displayed in a visual order. The user can correct a particular character to a different character. When recognized text needs to be sent to a separate application, an inverse bi-directional process is performed to convert the text from the visual order to the logical order.