Image Processing Parameterization for Machine Tool Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The parameterization of image processing systems for monitoring machine tools is complex due to numerous influencing parameters, leading to challenges in detecting subtle errors and optimizing process control without causing false alarms or overlooking faults.
Innovation Solution
A method that initializes the image processing system with preliminary parameters, acquires raw data images with good/bad case information, and iteratively adjusts parameters to optimize evaluation, allowing offline optimization without machine occupation, thereby accelerating parameterization and reducing unnecessary rejects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the image processing system uses multiple parameters for image evaluation, then the measurement precision and reliability of error detection is improved, but the device complexity and difficulty of parameterization increases
Solution Approach 1:
The patent applies preliminary action by automatically determining optimal parameter values before actual image evaluation begins. The system performs parameter optimization in advance using training images, storing the determined parameter values for subsequent use. This eliminates the need for manual parameter tuning and reduces parameterization complexity while maintaining high measurement precision through optimized parameters.
Solution Approach 2:
The system applies self-service by automatically selecting and optimizing parameters without external intervention. The image processing system autonomously determines optimal parameter values through iterative evaluation against training images, and automatically stores these parameters for future operations. This self-optimizing capability reduces the burden on operators while ensuring precise error detection.
2Reliability
If the image processing system is parameterized manually with careful selection of parameters, then the reliability of process control is improved, but the time required for parameterization and loss of time increases
Solution Approach 1:
The patent replaces manual parameter selection (mechanical/systematic approach) with an automated computational system. The system uses algorithms to automatically determine optimal parameter values by evaluating training images and iteratively optimizing parameters. This substitution dramatically reduces parameterization time while maintaining or improving process control reliability through consistent, objective parameter selection.
Solution Approach 2:
The system applies parameter changes by automatically adjusting and optimizing parameter values based on training data. Instead of manual parameter setting, the system iteratively modifies parameters to maximize evaluation accuracy, then stores the optimized values. This automated parameter optimization achieves high reliability while minimizing the time required for parameterization.
3Measurement precision
If the image processing system uses complex parameter optimization, then the measurement precision of subtle error detection is improved, but the device complexity and difficulty of operation increases
Solution Approach 1:
The patent extracts the complex parameter optimization process from the user's responsibility and isolates it within the automated system. The system independently performs iterative parameter optimization using training images, separating the complex computational task from simple user operations. Users only need to provide training images and initiate the process, while the system handles the complex parameter tuning automatically, maintaining high measurement precision for subtle error detection.
4Productivity
If the image processing system is optimized offline without machine occupation, then the productivity and loss of time are improved, but the measurement precision may be affected without real-time feedback
Solution Approach 1:
The patent uses copying by creating a set of training images that replicate various machining conditions, including both normal and defective states. These training images serve as proxies for real-time machining scenarios, allowing the system to optimize parameters offline without occupying the machine. The copied training data enables accurate parameter determination while maintaining measurement precision through diverse representative samples.
Data Source
Figure 1~2
Figure 3
AI summary
A method for parameterizing an image processing system is described, which uses a camera (K) to monitor the machining area of a machine tool. Images of a workpiece (W) and its surroundings are captured, and the image processing system (BV) then decides whether an image shows a pass or a miss. The following steps are used to parameterize the image processing system: In an initialization step (10), the image processing system is initialized with preliminary parameters. In a raw data acquisition step (20), a set of several images is saved as raw data independent of the image processing system's parameterization. In a parameter modification step (40), parameters of the image processing system are changed.In an assessment step (30), the raw data are processed by the image processing system using the parameters and classified into good cases and bad cases, and the classification into good cases and bad cases made by the image processing system is compared with the good/bad information stored together with the raw data.