Image Processing Device Character String Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing methods face difficulties in accurately detecting a character string from an image using simple operation inputs, especially when multiple character strings are present, requiring users to manually designate their target string.
Innovation Solution
An image processing device with a receiving unit, specifying unit, and detecting unit that receives input information, specifies the position of the input, and detects character strings with a smaller distance to the specified position, allowing for automatic selection of the target character string.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all detection fields indicating character strings are detected from an image, then the completeness of character string detection is improved, but the complexity of user operation increases when multiple character strings are present
Solution Approach 1:
The patent segments the detection process into two stages: first detecting all potential character string regions in the image, then selectively identifying the target character string based on user input position. This segmentation allows complete detection while simplifying user interaction by automatically selecting the relevant string.
Solution Approach 2:
The patent introduces an intermediary mechanism (user input position detection unit) that mediates between the user's simple input and the character string detection process. This intermediary automatically determines which detected character string is the user's target based on the input position, eliminating the need for users to manually select from multiple strings.
2Measurement precision
If conventional detection methods are used to detect all character strings in an image, then detection completeness is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary detection of all character string regions in the image first, storing their positions and information. Then, based on user input position, it quickly identifies the target character string without needing to re-process the entire image, significantly reducing processing time for the final selection.
Solution Approach 2:
The patent performs excessive detection by detecting all character strings in the image rather than just the target string. This excessive action is justified because it enables faster subsequent selection based on user input position, trading initial processing effort for faster final result delivery.
3Measurement precision
If manual designation is required to select target character string from multiple detection fields, then detection accuracy is improved, but ease of operation deteriorates
Solution Approach 1:
The patent implements self-service by enabling the system to automatically identify the user's target character string based on the user's input position. The system serves itself by using the user's natural input as the selection criterion, eliminating the need for additional manual designation steps while maintaining accurate target identification.
Data Source
AI summary
According to an embodiment, an image processing device includes a memory, and one or more hardware processors configured to function as a receiving unit, a specifying unit, and a detecting unit. The receiving unit receives input information input to an image. The specifying unit specifies the position of the input information. The detecting unit detects a character string having a smaller distance to the position than another character string, from the image.