Image Processor Automatic Reading Mode Determination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing technologies face challenges in accurately determining the reading mode for documents, leading to reduced reading speed and file size issues, especially when distinguishing between color, gray, and binary modes, and struggle with high accuracy at low resolutions.
Innovation Solution
An image processor that automatically determines the reading mode by performing color and gray detection processes, removing backgrounds and line segments, and converting image data based on detected information, allowing for high-quality and high-speed reading without the need for pre-scans and enabling efficient file compression and division.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pre-scan is performed to determine image type before main scan, then reading mode determination accuracy is improved, but reading speed is reduced due to requiring two reads
Solution Approach 1:
The patent applies preliminary action by performing background removal and line/character detection on the original image data before gray determination. This allows the system to use the full-resolution image for accurate structure detection, then use the processed image (with background and lines removed) for gray area determination, eliminating the need for a separate pre-scan while maintaining high accuracy.
2Extent of automation
If histogram is used for determining gray mode, then automation is improved, but determination accuracy deteriorates because characters also appear gray at low resolution
Solution Approach 1:
The patent applies the extraction principle by removing the background and line segments from the image data before performing gray determination. This extraction isolates the actual gray content (such as shaded portions in drawings) from elements that would otherwise be misidentified as gray (such as characters and background), enabling accurate automated determination without confusion from low-resolution character detection.
3Quantity of substance
If monochrome document is read in binary mode, then file size is reduced, but picture portions and shaded portions are not reproduced
Solution Approach 1:
The patent applies local quality by determining the reading mode (binary, gray, or color) for different regions of the document based on local characteristics. The system divides the document into multiple areas and applies appropriate processing to each area - binary for text-heavy regions, gray for shaded portions and drawings, and color for彩色 images - thereby optimizing both file size and reproduction quality for each local region.
4Manufacturing precision
If monochrome document is read in color mode, then image quality is improved, but file size becomes large due to unnecessary color components
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the color depth and format parameters based on the detected document characteristics. The system changes parameters such as color mode (color/gray/binary), resolution, and compression settings according to the determined reading mode, thereby optimizing the balance between image quality and file size for each specific document type.
Data Source
AI summary
An image processor comprising a control unit, wherein the control unit includes a determining unit that determines a reading mode suitable for a document based on image data corresponding to the document read by an reader; and a converting unit that converts the image data according to the reading mode. The determining unit includes a color detecting unit that detects color information based on the image data; a color determining unit that determines whether the reading mode is a color mode based on the color information; a background detecting unit that detects a background of the document; a background removing unit that removes the background; a gray detecting unit that detects a size of an area of a portion which is not removed in the image data as gray information; and a gray determining unit that determines whether the reading mode is a gray mode based on the gray information.