Image Anomaly Inspection With Corrective-Action Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing maintenance systems for image forming apparatuses do not effectively specify parts to be replaced based on image anomalies and lack a mechanism for inferring faulty portions and processing content corresponding to image anomalies and corrective actions.
Innovation Solution
An image inspection apparatus that collects information on corrective actions performed by users for image anomalies, trains a learning model using this data, and estimates corrective action content, enabling efficient part replacement and maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a maintenance system specifies workers and visiting routes based on part replacement requests, then worker efficiency is improved, but the system cannot identify which parts to replace from image anomalies
Solution Approach 1:
The patent replaces manual part identification and maintenance decision-making with an automated image analysis system. The image analysis apparatus automatically detects image anomalies, identifies faulty parts, and determines replacement decisions, substituting the mechanical/manual process with an automated optical and computational system.
Solution Approach 2:
The patent introduces an image analysis apparatus as an intermediary between the image forming apparatus and the maintenance worker. This intermediary automatically analyzes images, identifies anomalies, determines faulty parts, and provides maintenance decisions, bridging the gap between raw image data and actionable maintenance information.
2Measurement precision
If an image forming apparatus displays anomalous images and acquires user selection, then faulty portions can be identified, but the system does not record or collect data for inferring faulty portions from image anomalies
Solution Approach 1:
The patent implements a feedback mechanism where the image forming apparatus transmits both the anomalous image and the user's corrective action (part replacement information) back to the image analysis apparatus. This feedback loop allows the system to learn from actual maintenance actions and improve future fault identification accuracy.
Solution Approach 2:
The patent performs preliminary data collection and association of image anomalies with corrective actions. By recording and storing these paired data points before they are needed for inference, the system prepares a knowledge base that enables future automatic fault identification without requiring manual intervention for each new anomaly.
3Adaptability or versatility
If cloud computing is used for distributed data processing, then computing resources are efficiently utilized, but the system lacks a mechanism to associate image anomalies with corrective actions for learning model training
Solution Approach 1:
The patent creates a universal data collection and association mechanism that works across multiple image forming apparatuses connected to the image analysis apparatus. The system universally applies the same image anomaly detection and corrective action recording process to all connected devices, enabling aggregated learning model training across the entire network.
Data Source
AI summary
An image inspection apparatus that acquires information on a corrective action performed by a user for an image anomaly, and performs training of a learning model to estimate a candidate for corrective action content from an image anomaly using the image anomaly and the acquired information on the corrective action as learning data.