Image Processing Apparatus for Accurate Symbol Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies face challenges in accurately extracting images from original documents due to erroneous recognition of pre-formed symbols as management symbols, leading to incorrect association and management of extracted images.
Innovation Solution
An image processing apparatus that includes a reading unit, an acquiring unit, a color information extracting unit, a storage unit, an image extracting unit, and an image generating unit, which extracts images from regions surrounded by specific color information and generates new images based on layout position information associated with this information, using a multifunction peripheral with a CPU or MPU for control and storage of image data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a symbol previously formed on the original document is recognized as the management symbol, then the image extraction process can proceed, but erroneous recognition occurs leading to incorrect association and management of extracted images
Solution Approach 1:
The patent segments the image processing task into distinct phases: detection phase (identifying all symbols including hand-drawn markers and pre-formed symbols), verification phase (distinguishing hand-drawn markers from pre-formed symbols through color analysis), and extraction phase (extracting images associated with verified hand-drawn markers). This segmentation allows the system to maintain high productivity while ensuring accurate distinction between different symbol types to prevent erroneous recognition.
Solution Approach 2:
The patent introduces color information as an intermediary criterion to mediate between the detection of symbols and the final association of extracted images. By analyzing color properties (hand-drawn markers typically have distinct colors compared to pre-formed symbols), the system creates an intermediate verification step that resolves the contradiction between efficient detection and accurate recognition, ensuring that only properly identified hand-drawn markers trigger image extraction and association.
2Measurement precision
If all symbols on the original document are treated as potential management symbols, then no images are missed, but false positives increase leading to erroneous recognition
Solution Approach 1:
The patent applies local quality analysis by examining specific properties of each detected symbol rather than treating all symbols uniformly. The system analyzes local color characteristics, line thickness, and pattern features of each symbol to determine whether it is a hand-drawn marker or a pre-formed symbol. This localized analysis enables the system to maintain high detection completeness while minimizing false positives by making recognition decisions based on specific local attributes of each symbol.
Solution Approach 2:
The patent changes the recognition parameter from simple symbol presence detection to multi-parameter analysis including color information, line width, and spatial distribution. By introducing color as an additional parameter and analyzing its characteristics (hand-drawn markers typically exhibit different color properties than pre-formed symbols), the system transforms the recognition process to achieve both high completeness and low false positive rates through comprehensive parameter evaluation.
Data Source
AI summary
An image processing apparatus includes a reading unit, an acquiring unit, a color information extracting unit, a storage unit, an image extracting unit, and an image generating unit. The reading unit reads an image on a sheet. The acquiring unit acquires image data from the image read by the reading unit. The color information extracting unit extracts color information from the image data. The storage unit stores specific color information. The image extracting unit extracts second image data surrounded by a line with the specific color information in first image data that the acquiring unit acquires from an image on a first sheet which is read by the reading unit. The image generating unit generates third image data based on the second image data.