Editing Table Regions in Scanned Images Using Cell Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image editing systems for scan images face challenges in accurately determining the position of table regions, leading to incorrect deletion of intended images and improper editing.
Innovation Solution
An information processing apparatus that performs binarization, region division, and cell recognition to identify table regions, allowing for precise selection and editing of cells within the image, with features like recognized cell position change handles and post-edit text arrangement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automatic table region identification is performed using existing image editing systems, then editing efficiency is improved, but position determination accuracy deteriorates leading to incorrect deletion of images
Solution Approach 1:
The patent segments the table region identification process into multiple independent steps: binarization of the scan image, region division to identify candidate areas, and cell recognition to determine precise table boundaries. This segmentation allows each step to be optimized independently, improving both efficiency and accuracy compared to monolithic automatic identification systems.
Solution Approach 2:
The patent performs preliminary binarization and region division before final cell recognition and editing operations. By pre-processing the image to identify candidate regions and eliminate non-table areas, the system prepares accurate position data in advance, preventing incorrect deletions during the actual editing phase while maintaining high throughput.
2Adaptability or versatility
If table region identification is performed to enable editing, then editing capability is improved, but reliability deteriorates due to incorrect position determination
Solution Approach 1:
The patent implements feedback mechanisms where the system continuously refines table region identification based on recognition results. The cell recognition process uses feedback from region division to adjust and confirm precise table boundaries, ensuring reliable position determination before editing operations are executed, thus preventing incorrect modifications.
3Measurement precision
If multiple processing steps (binarization, region division, cell recognition) are performed, then identification accuracy is improved, but device complexity increases
Solution Approach 1:
The patent divides the complex identification process into three manageable segments: binarization, region division, and cell recognition. Each segment performs a specific function with well-defined inputs and outputs, making the overall complex system easier to implement, debug, and maintain compared to a single monolithic processing algorithm.
Solution Approach 2:
The patent performs preliminary binarization and region division that simplify the subsequent cell recognition step. By pre-processing the image to create binary masks and identify candidate regions, the system reduces the complexity of the final recognition stage, as it only needs to operate within pre-defined boundaries rather than analyzing the entire image.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
There is provided an information processing apparatus. When an information processing apparatus receives a selection of a position under a situation where an image including a plurality of recognized cells is displayed on a display unit, the information processing apparatus displays an editing region for allowing a user to edit a text included in a recognized cell including the position, and a handle for changing the position of the recognized cell including the position which has been received a selection.