Mobile Terminal Partial Region Text Translation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current mobile terminal translation applications cannot accurately translate text within a partial region of an image, requiring translation of all text in the image, which lowers recognition rates.
Innovation Solution
A mobile terminal with a camera and display unit that allows users to select a region of interest on the screen, activating a translation mode and displaying a GUI window for text recognition within that region, enabling translation of specific text areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system translates all text in the entire image, then complete translation coverage is achieved, but text recognition rate decreases due to processing large areas with potential noise
Solution Approach 1:
The patent divides the image processing into two stages: first, the entire image is scanned to detect text regions and generate candidate bounding boxes; second, the user selects specific candidate regions for translation. This segmentation allows the system to process the whole image for detection while limiting translation to only the selected regions, improving recognition accuracy by focusing on relevant areas.
Solution Approach 2:
The patent implements local quality by allowing different processing intensities for different regions. The initial text detection is performed globally across the entire image, but the translation operation is applied locally only to user-selected regions. This ensures high recognition rates for selected text while avoiding noise from unrelated areas of the image.
2Adaptability or versatility
If the system processes the entire image for text detection, then all text can be potentially translated, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary text detection and candidate region generation across the entire image before the user makes selections. This preliminary action prepares the data structure and identifies potential text regions in advance, so that when the user selects specific regions for translation, the system can quickly process only those selected areas without re-scanning the entire image.
Solution Approach 2:
The patent implements partial action by performing full-image text detection (excessive action for completeness) but then limiting the translation operation to only the user-selected regions (partial action for efficiency). The system does more detection than strictly necessary for the final translation, but this enables faster and more targeted processing of the actual translation task.
3Measurement precision
If the system requires selecting partial regions from the image, then translation precision is improved, but ease of operation decreases due to additional selection steps
Solution Approach 1:
The patent implements self-service by automatically generating candidate text regions and presenting them to the user for selection. Instead of requiring the user to manually draw or specify text regions, the system performs the tedious work of detecting and proposing text boundaries automatically, then allows the user to simply select from these pre-generated candidates. This maintains translation precision while significantly improving ease of operation.
Data Source
AI summary
A mobile terminal and controlling method thereof are provided. The mobile terminal includes a camera, a display unit configured to display a preview image obtained by the camera, and a controller configured to detect a first command for entering a translation mode, output a GUI window for selecting at least one partial region of a text included in the preview image to the display unit in response to the detected first command, control the display unit to display a translation result of a text corresponding the GUI window on the preview image.