Image Reading Through Transparent Folders With Inner Edge Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image reading apparatuses struggle to efficiently cut out individual images of documents when using a transparent document folder, as they mistakenly cut out images based on the folder's edges rather than the document edges, requiring documents to be placed one by one on the glass plate.
Innovation Solution
An image reading apparatus with a transparent member and a processor that detects document edges within a holding member's transparent portion, allowing documents to be placed in a folder visible through the transparent portion, and cuts out images based on detected document edges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a transparent document folder holding multiple documents is used for reading, then the documents can be read collectively without placing them one by one, but the image reading apparatus cannot cut out individual document images because it detects the folder's edge instead of the document edges
Solution Approach 1:
The patent applies local quality by differentiating the handling of edges based on their location and characteristics. The processor detects multiple edges in the read image and identifies which edges correspond to the transparent document folder versus which correspond to individual documents. By analyzing edge properties (such as continuity, position, and relationship with detected documents), the system locally adapts its processing to distinguish between the folder boundary and document boundaries, enabling accurate extraction of individual document images even when they are contained within a transparent folder.
2Measurement precision
If multiple documents are placed on the document positioning glass plate one by one, then individual document edges can be accurately detected and images can be cut out separately, but it takes significant time and effort
Solution Approach 1:
The patent implements preliminary action by first reading multiple documents simultaneously while they are contained in the transparent document folder, before the actual image extraction is performed. The system pre-processes the read image by detecting all edges, identifying document regions, and preparing extraction parameters in advance. This allows the user to simply place the folder with multiple documents on the glass plate once, and the system automatically handles the complex edge detection and image separation tasks, significantly reducing the time required compared to placing documents one by one.
3Adaptability or versatility
If the transparent document folder is used for reading, then documents can be read as a group, but the edge detection algorithm mistakenly identifies the folder edge as the document edge, causing all documents to be output as a single image
Solution Approach 1:
The patent applies segmentation by dividing the edge detection and image extraction process into distinct stages. First, the system detects all edges in the read image and segments them into potential document edges and potential folder edges. Then, using the detected document regions and edge characteristics, the system segments the transparent folder area from the individual document areas. This multi-stage segmentation allows the system to maintain the flexibility of reading multiple documents together in a folder while achieving accurate extraction of individual document images by distinguishing between the folder boundary and document boundaries.
Data Source
AI summary
An image reading apparatus includes a transparent member having a placement surface on which a document is to be placed, a reading unit configured to read an image of the document placed on the transparent member via the transparent member, and a reading mode in which an image of a document, placed in a holding member with a transparent portion, is read. The document to be read is placed in the holding member so as to be visible via the transparent portion and the holding member is placed on the placement surface with the transparent portion of the holding member in contact with the placement surface. A processor detects an edge in an image acquired by the reading unit in the reading mode, and to extract an image surrounded by an edge detected in an area surrounded by an edge corresponding to the holding member in the reading mode.