Wood Shingle Machine Vision Classification for False Defect Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The shingle manufacturing industry faces challenges in accurately classifying wood shingles using machine vision due to the subjective nature of defect identification and the difficulty in distinguishing between real and false defects, which limits the reliability of robotic machinery and increases the need for human expertise.
Innovation Solution
A machine vision system that classifies wood shingles by comparing images of defects to a database of confirmed defects, using a 'Skip-a-Scan' method to predict subsequent shingle quality, and incorporating artificial intelligence and human subjectivity to improve accuracy, while preventing false defects from being added to the database, and employing a 'Clear-or-Better' approach to simplify grading.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine vision is used to classify wood shingles, then productivity is improved, but measurement precision deteriorates due to difficulty in distinguishing real defects from false defects
Solution Approach 1:
A database of confirmed defects serves as an intermediary between the machine vision system and the classification decision. The system compares detected defects against this curated database to determine whether they represent actual defects or false positives, thereby improving measurement precision while maintaining automated classification speed
Solution Approach 2:
The database of confirmed defects is prepared in advance through manual verification and expert review. This preliminary action creates a reference library that the machine vision system can query during automated classification, enabling accurate defect identification without requiring real-time expert intervention
2Measurement precision
If comprehensive defect criteria are used to improve classification accuracy, then measurement precision is improved, but device complexity increases due to the number of defect types and grading rules
Solution Approach 1:
The complex classification task is segmented into two independent components: (1) defect detection and characterization, and (2) classification decision-making based on defect properties. The database stores defect characteristics separately from grading rules, allowing each component to be optimized independently without increasing overall system complexity
3Measurement precision
If manual verification of defects is performed to improve measurement precision, then classification accuracy is improved, but loss of time increases
Solution Approach 1:
Instead of manually verifying every detected defect, the system applies partial manual verification only to cases where the machine vision system has low confidence or detects ambiguous features. The database of confirmed defects provides pre-verified reference cases that eliminate the need for manual verification in clear-cut situations, reducing time loss while maintaining precision
Data Source
AI summary
A method of wood shingle classification and sawing using machine vision comprises the steps of taking an image of a wood slab in a wood block and identifying a defect in that slab; comparing an image of this defect to images of confirmed defects in a database of confirmed defects to find a match of the defect in these images. If a match is not found, sawing a shingle from the slab and classifying the shingle while making abstraction of the defect. In a second aspect, when images of two consecutive shingles are identical, a third and subsequent shingles can be sawn from a block without taking images thereof. In another aspect, the comparing of images is done by an artificial intelligence system that is trained on a database of images that are associable to the subjectivity of experienced shingle sawyers.