Image Recognition Controller for Text Extraction in Portable Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current portable devices with electronic dictionary functions require users to manually input text for translation, leading to inefficiency and divided attention, as they cannot effectively extract and output relevant information from images.
Innovation Solution
A method and apparatus that utilize image recognition to identify effective information within an image, allowing users to recognize objects by indicating their position with a second object, thereby extracting and outputting related information without manual input, such as using a camera unit, display unit, and image recognition controller to recognize position information and extract relevant data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual text input is used for translation in electronic dictionary function, then translation capability is provided, but operation time increases and user attention is divided
Solution Approach 1:
The patent replaces manual text input (mechanical typing operation) with image recognition technology. The camera captures an image containing text, the image recognition controller automatically extracts and translates the text, eliminating the need for manual typing while maintaining translation functionality.
Solution Approach 2:
The system performs automatic text extraction and translation without requiring user intervention for text input. The image recognition controller autonomously processes the captured image, identifies text regions, extracts text content, and retrieves translations, making the system serve itself rather than requiring continuous user input.
2Adaptability or versatility
If OCR technique is used to analyze image information, then translation function is provided, but effective information recognition difficulty increases and user attention is divided
Solution Approach 1:
The patent segments the image processing task into distinct stages: the image recognition controller first identifies text regions within the captured image, then extracts text content from those regions, and finally retrieves translations. This segmentation simplifies the overall process by breaking down complex image analysis into manageable steps, reducing recognition difficulty.
Solution Approach 2:
The image recognition controller acts as an intermediary between the camera and the translation database. It processes the raw image data, extracts meaningful text information, and queries the translation database, serving as a mediator that simplifies the interaction between image input and translation output.
3Productivity
If user divides attention between dictionary and textbook, then translation can be performed, but concentration difficulty increases
Solution Approach 1:
The patent merges the textbook viewing and translation functions into a single integrated process. The camera captures the textbook content, the image recognition controller extracts and translates the text, and the translation result is displayed alongside or near the original content, allowing users to view both the source material and its translation simultaneously without switching between applications.
Data Source
AI summary
A method for recognizing an image and an apparatus using the same. The method includes: receiving image information including a first object and a second object; recognizing position information of the first object indicated by the second information in the received image information; extracting effective information included in the first object of the received image information in response to the recognized position information; and outputting related information corresponding to the recognized effective information.