Article Estimation System Using Spine Text Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems cannot identify books or similar articles from spine images due to the lack of spine images in databases, requiring users to photograph covers, which is inconvenient and limits the ability to identify multiple books at once.
Innovation Solution
An article estimating system that acquires images, extracts image areas, analyzes them using OCR, and uses additional information from reference articles to narrow down identification candidates, allowing identification from spine images by associating size, page number, series, and genre information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image matching process is used to identify books, then identification accuracy is improved, but spine images are not stored in database so identification cannot be performed
Solution Approach 1:
The system changes the approach from image-based matching to parameter-based identification. Instead of comparing spine images, the system extracts text parameters (title, author, publisher) from the spine image using OCR, then queries the database using these text parameters. This parameter transformation enables spine image identification without requiring stored spine images in the database.
2Reliability
If user photographs book cover to identify the book, then identification is possible, but user effort increases and multiple books cannot be photographed efficiently
Solution Approach 1:
The system enables self-service identification where the book itself provides the identification information through its spine text. The OCR technology automatically extracts text from the spine image, and the system automatically matches this text with database records. This eliminates the need for user intervention to manually input book details or select from multiple options, making the process as easy as taking a photograph.
3Adaptability or versatility
If only cover images are stored in database, then image matching can be performed, but spine images cannot be identified even when photographed
Solution Approach 1:
The system segments the identification process into two independent parts: (1) text extraction from the spine image using OCR, and (2) database query using the extracted text parameters. This segmentation allows the system to identify books by spine text without requiring spine images to be stored in the database, thereby maintaining identification flexibility while avoiding the need to store and manage spine image data.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables efficient identification of articles like books from spine images without the need for image matching, reducing user effort and processing load, and allowing multiple books to be identified simultaneously.
Implementation Method 1
an analysis means which analyzes the plurality of extracted image areas to acquire analysis information
Data Source
AI summary
A server 2 includes an extraction unit 21, an analysis unit 22, a first estimating unit 24, an information acquisition unit 25 and a second estimating unit 26. The extraction unit 21 extracts an image area for each article. The analysis unit 22 analyzes the image area to acquire analysis information. The first estimating unit 24 narrows down candidates estimated to correspond to the article in the image area based on the analysis information. When the candidates were able to be narrowed down, the information acquisition unit 25 acquires additional information additional information of a reference article. The second estimating unit 26 attempts a narrowing process based on the additional information of the reference article in addition to the analysis information, for the image area including a spine, which is an image area in which candidates were unable to be narrowed down.