Toner Classification via Printout Image Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Low-quality, non-OEM toner can cause damage to multifunctional printers by leaking into the inner mechanisms and producing subpar print quality, leading to equipment damage and increased service calls due to incompatibility issues that are not immediately visible.
Innovation Solution
A deep neural network using machine learning analyzes printout images to differentiate between OEM and non-OEM toner, employing a convolutional neural network for image classification to predict toner quality and identify potential damage risks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If non-OEM toner is used to reduce costs, then manufacturing cost is reduced, but printer reliability deteriorates due to potential damage from low-quality toner
Solution Approach 1:
The system performs preliminary classification of toner cartridges before they cause damage to the printer. By analyzing printout images and identifying non-OEM toner early in the printing process, the system can alert users or automatically adjust settings to prevent the toner from damaging the printer's internal mechanisms, thus maintaining reliability while allowing cost-saving toner usage.
2Ease of manufacture
If non-OEM toner is used to reduce costs, then manufacturing cost is reduced, but print quality deteriorates due to lower quality output
Solution Approach 1:
The system implements a feedback mechanism where printout images are continuously analyzed and classified to identify non-OEM toner usage. When non-OEM toner is detected, the system can provide feedback to the user about the reduced print quality, suggest optimizations to maintain acceptable quality standards, or alert users to replace the toner when quality thresholds are not met, thus managing quality expectations while allowing cost-saving toner usage.
3Reliability
If automated toner classification is implemented to protect printer reliability, then printer reliability is improved, but device complexity increases due to additional analysis systems
Solution Approach 1:
Instead of adding complex physical sensors or analysis equipment to the printer itself, the system creates a digital copy of the printout image and performs the toner classification analysis on this copy. This approach protects printer reliability by identifying problematic toner through image analysis while avoiding the need for complex additional hardware in the printer, thus minimizing the increase in device complexity.
Data Source
AI summary
A system and method for toner classification from device printouts includes applying machine learning to electronic documents formed from scanned printout images. A training set is formed by scanning documents known to be printed with OEM toner, supplemented by scanning documents known to be printed with non-OEM toner. When new printouts are made, they are scanned and analyzed by an AI/ML server and classified as printed by OEM toner or by non-OEM toner.