Wafer Box Clip State Detection Using Background Masking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting clip integrity in manufacturing processes, particularly in wafer production, face challenges due to small or hard-to-see latches, positioning errors, and interference from background regions, leading to incorrect AI detection results.
Innovation Solution
A method that determines a desired region for analysis in a detection image, masks the background region to improve clip recognition, and uses a trained neural network model to differentiate between sealed and unsealed clips based on masked images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If AI model is used to detect clip state, then detection cost is reduced and detection accuracy is improved, but positioning errors occur when clip is out of preset location by large distance
Solution Approach 1:
The patent divides the detection process into multiple stages: first detecting clip position, then selectively masking background regions based on position accuracy. This segmentation allows the system to handle positioning errors by separating the position detection function from the state detection function, improving overall reliability while maintaining accuracy.
Solution Approach 2:
The patent introduces an intermediary masking mechanism that filters out background regions before state detection. This intermediary step acts as a mediator between position detection and state detection, preventing background interference from causing incorrect results when clips are slightly out of position.
2Loss of information
If background region is included in detection image, then complete clip information is captured, but background interference causes incorrect detection results
Solution Approach 1:
The patent extracts and removes background regions from the detection image based on clip position information. By taking out the interfering background elements while preserving the clip region, the system maintains information completeness for the clip while eliminating background interference that would cause detection errors.
Solution Approach 2:
The patent applies different processing quality to different regions: the clip region is processed with full detail for state detection, while the background region is masked or reduced in quality. This local quality differentiation ensures high detection accuracy for the clip without being compromised by background interference.
3Ease of operation
If fixture provides assistance to clip position, then clip positioning is facilitated, but close tolerance is not achieved and positioning errors occur
Solution Approach 1:
The patent implements a feedback mechanism where the detected clip position is used to adjust the detection strategy. When position errors are detected, the system feeds back to modify the masking parameters and detection region, allowing the system to compensate for fixture limitations and achieve accurate state detection despite imperfect positioning.
Data Source
AI summary
A method for detecting from images correct placement and function, or incorrect placement and function, of a clip of a transportation box of wafers in sterile or similar conditions obtains an image template comprising features of clip and obtains a first detection image of a working clip. An object region focusing on the imaged clip in the first detection image is determined according to the image template. Part of the working image is selected as a first preset location. The part of the image is masked to obtain a second detection image, the masking obscures the background region of the part of the image but not the clip-object region, and displays the unobscured clip-object region. The second detection image is input into a trained neural network model to determine whether the clip is in sealed or unsealed state. An electronic device and a non-transitory storage medium are also disclosed.