Image Processing System Prohibited Content Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information processing systems lack the capability to effectively detect and handle image information containing prohibited content, such as specific characters or figures, which can lead to unauthorized transmission or disclosure.
Innovation Solution
An information processing system with a registration unit for storing prohibited information associated with attribute information and a controller that performs error handling by identifying and managing prohibited content within received image information, including options like discarding, replacing, or reporting such content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image recognition is performed on all received image information, then detection precision of prohibited content is improved, but processing time and system complexity increase
Solution Approach 1:
The image information processing is divided into multiple stages: first extracting attribute information from image metadata, then performing image recognition only when attribute information indicates potential prohibited content. This segmentation reduces overall processing time while maintaining detection precision for prohibited content.
Solution Approach 2:
Attribute information extraction is performed as a preliminary action before full image recognition. By checking attribute information first (such as file type, size, or metadata indicators), the system can quickly filter out safe images and only apply comprehensive image recognition to suspicious cases, thereby reducing total processing time.
2Measurement precision
If comprehensive image recognition is performed on all image information, then detection precision is improved, but device complexity increases
Solution Approach 1:
The detection system is segmented into multiple components: an attribute information extraction module that handles metadata analysis, and an image recognition module that performs comprehensive analysis only when needed. This modular segmentation reduces overall system complexity while maintaining high detection precision.
Solution Approach 2:
Attribute information serves as an intermediary between the received image information and the full image recognition process. By using attribute information as a filtering intermediary, the system avoids the complexity of performing comprehensive image recognition on all images while maintaining detection precision for prohibited content.
3Reliability
If prohibited content detection is implemented, then data security is improved, but loss of information increases due to discarding or replacing valid image data
Solution Approach 1:
The system extracts only the prohibited content portions from image information rather than discarding entire images. By taking out and identifying specific prohibited elements within images, the system maintains data security while preserving the majority of valid image information for legitimate use.
Solution Approach 2:
The detection and handling of prohibited content is applied locally to specific regions or portions of images rather than uniformly to entire images. This allows the system to maintain high data security for prohibited content while preserving the quality and usability of non-prohibited portions of the same image.
Data Source
AI summary
An information processing system includes a registration unit and a controller. The registration unit registers prohibited information associated with at least one of pieces of attribute information on a per piece-of-attribute information basis. The controller refers to the pieces of attribute information and the prohibited information. The controller performs error handling if received image information includes the prohibited information associated with a piece of attribute information assigned to the image information.