Image Processing OCR for Partial Personal Name Anonymization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing apparatuses struggle to effectively anonymize personal information in documents, particularly personal names, while maintaining readability for authorized users.
Innovation Solution
An image processing apparatus with an OCR process to extract personal names, recognize target regions, and perform anonymization by making non-initial characters in personal names unrecognizable, using methods such as blacking out, Romanization, or replacing with initial characters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If complete anonymization of personal names is performed, then privacy protection is improved, but readability and usability for authorized users deteriorates
Solution Approach 1:
The patent applies different levels of anonymization to different parts of the personal name. Specifically, the first character (family name initial) is preserved while subsequent characters are anonymized. This creates a localized quality distinction where the initial character maintains readability for user identification, while the remaining characters provide sufficient privacy protection.
2Object-affected harmful factors
If personal information is completely concealed, then security is improved, but the ability to identify documents deteriorates
Solution Approach 1:
The patent implements partial anonymization rather than complete concealment. By preserving the first character of personal names while anonymizing the rest, the system applies just enough anonymization to achieve security goals while retaining sufficient information for document identification and user recognition.
Data Source
AI summary
An image processing apparatus includes an image reading portion and a control portion that performs an anonymizing process on original image data to generate output image data in which personal information is anonymized. When generating the output image data, the control portion extracts text data by an OCR process on the original image data, extracts from the text data a personal name as personal information, recognizes as a target region a region of the original image data that contains the personal name, and performs a process of making unrecognizable any character other than an initial character in a character string constituting the personal name in the target region.