Assistive Indications for Obscured Text via Character Spelling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Visually impaired individuals face challenges in reading documents due to obscured text, which existing software solutions often describe unclearly, requiring improved methods for seamless and intuitive text reading assistance.
Innovation Solution
A computer-implemented method that uses optical character recognition to generate assistive indications, including a synthetic voice that recites recognized words and spells out unrecognized words character by character, along with visual and haptic cues to aid in understanding incomplete words.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If existing software applications detect text and describe contents to visually impaired individuals, then text accessibility is improved, but clarity of description deteriorates when text is obscured
Solution Approach 1:
The system segments the text recognition process into two distinct stages: first recognizing complete words, then identifying individual characters. This segmentation allows the system to handle obscured text differently - recognizing complete words when visible and spelling out characters when obscured, thereby maintaining both accessibility and clarity in the description.
Solution Approach 2:
Instead of attempting to recognize complete words first and then dealing with obscured portions, the system inverts the approach by first identifying all visible complete words and then separately processing individual characters. This inversion allows the system to provide clear descriptions by distinguishing between recognized words and characters that need to be spelled out, improving overall description clarity.
2Ease of operation
If software applications provide seamless text conveyance with little user interaction, then ease of operation is improved, but adaptability to obscured text deteriorates
Solution Approach 1:
The system performs self-service by automatically detecting which words are completely visible and which characters are obscured, then automatically adjusting its output accordingly. It seamlessly switches between reading complete words and spelling out characters without requiring user intervention, thereby maintaining ease of operation while adapting to various text visibility conditions.
Solution Approach 2:
The system changes its operational parameters dynamically based on text visibility conditions. When text is fully visible, it uses word-level recognition parameters; when text is obscured, it switches to character-level spelling parameters. This parameter change allows the system to maintain seamless operation while adapting to different text conditions without requiring user interaction.
3Loss of information
If the system recites all words including unrecognized ones, then completeness of text conveyance is improved, but comprehension of incomplete words deteriorates
Solution Approach 1:
The system applies local quality by treating recognized words and unrecognized characters differently in its output. Complete words are read as whole units for efficient comprehension, while individual characters are spelled out separately when obscured. This differential treatment ensures that the system maintains completeness by including all text content while improving comprehension by adapting the delivery method to the specific local conditions of each word or character.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances the effectiveness of reading documents for visually impaired users by providing clear auditory, visual, and haptic indications, reducing user interaction and improving comprehension of incomplete words.
Implementation Method 1
generating, by the computing device, based at least in part on the image data and one or more optical character recognition operations, character data including one or more characters associated with the at least one image
Implementation Method 2
generating, by the computing device, one or more auditory indications that can include a synthetic voice that recites the one or more recognized words and the one or more unrecognized words
Data Source
AI summary
Methods, systems, devices, and tangible non-transitory computer readable media for generating assistive indications are provided. The disclosed technology can include accessing image data that includes at least one image. Character data can be generated based at least in part on the image data and one or more optical character recognition operations. Further, the character data can include one or more characters associated with the at least one image. One or more characters that are associated with one or more recognized words and the one or more characters that are associated with one or more unrecognized words can be determined based on the character data. One or more auditory indications including a synthetic voice reciting the one or more recognized words and the one or more unrecognized words can be generated. Furthermore, the synthetic voice can recite each of the one or more unrecognized words one character at a time.