Defect Ranking GUI for Optical Sorting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical sorting machines face challenges in optimizing the setup to effectively remove defects from bulk foodstuffs while minimizing the removal of good products, due to overlapping detection criteria and density-related separation issues, leading to uncertainty in adjusting sorting parameters and ejector blast areas.
Innovation Solution
A computerized system with a pattern recognition system and graphical user interface that processes image data to identify and rank defects according to multiple criteria, displaying thumbnails of defects and allowing operators to adjust sensitivity levels and ejection parameters, enabling precise control over the sorting process and visualization of defect detection and rejection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the sensitivity of the sorting criterion is increased to remove more defects, then the proportion of good product incorrectly classified as defect increases
Solution Approach 1:
The patent segments the defect detection process into multiple independent sorting criteria (e.g., colour, shape, size, texture) that operate simultaneously. Each criterion targets specific defect types with optimized sensitivity levels, allowing the system to detect different defect categories without cross-contamination. This segmentation enables selective rejection based on defect type while preserving good product.
Solution Approach 2:
The system dynamically adjusts sensitivity parameters for each sorting criterion based on the specific defect types being targeted. By changing detection parameters (thresholds, weightings) for different criteria, the system optimizes the balance between defect detection and good product preservation for each defect category, rather than using a single high sensitivity setting for all defects.
2Reliability
If multiple ejectors are fired simultaneously and/or the duration of gas blast is extended to ensure defective article rejection, then the area of intersection of gas with product stream increases causing removal of surrounding good product
Solution Approach 1:
The system applies local quality by directing gas blasts from multiple ejectors at precisely controlled locations corresponding to detected defect positions. Each ejector is activated selectively based on the spatial coordinates of detected defects, creating localized rejection zones that target only defective articles while minimizing gas interaction with surrounding good product.
Solution Approach 2:
The system dynamically adjusts the duration and intensity of gas blasts based on real-time defect detection data. The ejector activation timing and duration are optimized according to the position and velocity of detected defects, allowing precise control over the gas-product stream interaction area to reject defects while preserving good product.
3Reliability
If the area of intersection of gas blast with product stream is extended to account for position and velocity uncertainties, then the likelihood of removing good product increases
Solution Approach 1:
The system performs preliminary measurement and tracking of product article position and velocity before the rejection decision is made. By measuring these parameters in advance and using them to predict future positions, the system can calculate precise rejection zones that account for uncertainties without unnecessarily expanding the gas blast area, thus targeting only defective articles.
Solution Approach 2:
The system uses feedback from continuous position and velocity measurements to dynamically adjust the rejection zone calculations. By incorporating real-time feedback on article motion, the system optimizes the gas blast parameters (timing, duration, location) to match the actual trajectories of defective articles, minimizing good product removal while ensuring defect rejection.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Facilitates the optimization of sorting machine setup by providing operators with clear guidance on adjusting parameters, reducing the unintentional removal of good products and improving the accuracy of defect identification and rejection, thereby enhancing both quality and yield.
Implementation Method 1
a camera for generating image data of articles in the product stream at an imaging zone
Implementation Method 2
a sorting zone downstream of the imaging zone, and ejectors at the sorting zone for selectively ejecting articles from the stream
Data Source
AI summary
Inspection apparatus comprises a feed system for delivering a stream of articles to an imaging zone. A camera generates image data from the stream at the imaging zone for processing by a computer. The computer comprises a pattern recognition system for identifying defects in areas from the image data, and for ranking identified defects. The pattern recognition system is programmed to operate according to multiple defect criteria. The computer is also coupled to a graphical user interface to display the areas identified from the image data as thumbnails on the interface arranged according to rank of the identified defects in the areas, in each of at least two defect criteria. The areas from the generated image data will normally be defined around each identified defect with the defect central therein. These areas, or thumbnails, can overlap.