Video-Based Process State Evaluation Using Hue Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monitoring systems for manufacturing processes, such as those using monitoring cameras, burden operators with increased workload due to the need for real-time monitoring or post-processing video review, and require AI training, which also increases operator workload.
Innovation Solution
An information providing apparatus and method that calculates an evaluation value for the manufacturing process using conversion information and video data to assess the state of the process, allowing for automated alarm notifications and reducing the need for manual monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If real-time video monitoring is performed by operators, then the state of the monitoring target can be observed, but the burden on the operator increases
Solution Approach 1:
The patent replaces the mechanical system of manual video monitoring with an automated image processing system. The processor automatically extracts color information from video frames, converts it to hue values, and generates evaluation metrics without requiring operator intervention. This substitution eliminates the burden on operators while maintaining monitoring reliability through automated analysis of the manufacturing process state.
Solution Approach 2:
The monitoring system performs self-service by automatically analyzing video data without human intervention. The system extracts color information, converts it to hue values, and generates evaluation metrics autonomously. This self-service capability allows the system to monitor the manufacturing process state continuously without requiring operators to manually review video footage, thereby reducing operator burden while maintaining reliable monitoring.
2Reliability
If video review is performed after storage, then oversight risk is reduced, but the burden on the operator increases
Solution Approach 1:
The system performs preliminary action by automatically generating evaluation metrics from video data in real-time or near-real-time, before operators would need to review the footage. The processor continuously extracts color information and converts it to hue values, creating ready-to-analyze evaluation data. This preliminary processing eliminates the need for operators to spend time manually reviewing stored video, as the critical information has already been extracted and organized into evaluation metrics.
Solution Approach 2:
The system extracts only the essential information from video data - specifically color information and its conversion to hue values - separating this critical data from the complete video footage. By taking out only the necessary evaluation metrics rather than requiring operators to review entire video sequences, the system maintains monitoring accuracy while significantly reducing the time operators would need to spend on review.
3Extent of automation
If AI training is implemented for monitoring, then automated detection is achieved, but operator workload increases
Solution Approach 1:
The system uses parameter changes by converting color information into hue values through a defined conversion process. This parameter transformation creates a standardized metric that can be directly evaluated without requiring complex AI training. The conversion from raw color data to hue values provides a straightforward automated detection mechanism that reduces operator workload, as the system automatically processes and evaluates the transformed parameters without human intervention or training requirements.
Data Source
AI summary
An information providing apparatus includes a processor configured to receive conversion information including each conversion value corresponding to each hue, acquire video data of a monitoring target captured by an imaging device; and calculate an evaluation value for evaluating a state of the monitoring target on a basis of the received conversion information and the acquired video data.