Board Pack Image Retrieval Using Segmented Fingerprint Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In wood processing plants, the tracking of individual boards after storage becomes inefficient due to the high volume of data in system memory, leading to excessively long processing times, especially when identifiers on board packs become ruined or detached, and existing tracking systems based on 'fingerprint' features are computationally intensive.
Innovation Solution
A method involving a preparation step to generate and save digital datasets for known packs of boards, including reference images and features, and a retrieval step using image processing and machine learning algorithms to identify matches between boards, reducing the need to compare against the entire digital archive by focusing on sub-groups with common features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If tracking is based on fingerprint features of individual boards, then identification precision is improved, but processing time increases excessively due to high computational load
Solution Approach 1:
The patent segments the large digital archive into multiple sub-groups based on common features (wood type, dimensions, origin). Instead of comparing a board's fingerprint against all boards in the archive, the system first identifies the relevant sub-group and then performs matching only within that subset. This segmentation dramatically reduces the number of comparisons needed while maintaining identification precision.
Solution Approach 2:
The patent performs preliminary actions by pre-organizing the digital archive into structured sub-groups with common features before the actual tracking operation. This pre-processing creates an efficient search space structure that enables rapid retrieval and matching operations, avoiding the need to process the entire archive during real-time tracking.
2Extent of automation
If identifiers are applied to packs of boards, then tracking capability is improved, but reliability deteriorates when identifiers become ruined or detached
Solution Approach 1:
The patent creates a digital copy (fingerprint) of each board's surface features and stores it in the digital archive. This digital fingerprint serves as a permanent, unalterable record that does not deteriorate like physical identifiers. The system can retrieve and match this digital copy at any time, ensuring reliable identification even when physical identifiers are damaged or removed.
3Measurement precision
If all boards are stored in system memory for tracking, then identification accuracy is maintained, but computational complexity increases
Solution Approach 1:
The patent segments the comprehensive digital archive into organized sub-groups based on shared characteristics. This segmentation maintains access to all board data for accurate identification while structuring the data in a way that reduces computational complexity during matching operations by limiting the search space to relevant sub-groups only.
Solution Approach 2:
The patent performs preliminary organization of the digital archive into feature-based sub-groups before matching operations. This pre-structuring of data reduces the computational burden during actual tracking by eliminating the need to process unrelated boards, while still maintaining access to the complete board inventory for accurate identification.
Data Source
AI summary
Retrieving saved information relative to a first pack of boards, which belongs to a plurality of known packs of boards, entails a preparation step comprising a generation step and a saving step for saving reference features in a digital memory together with the information relative to the known pack of boards, creating a digital dataset relative to the known pack of boards, the set of the digital datasets constituting a first digital archive, and a retrieval step comprising an observation step in which at least one digital photograph of the first pack of boards is acquired, a comparison step for comparing first features linked to the digital photograph with the saved reference features, to identify a match with one of the saved reference images, and an extraction step for extracting from the first digital archive information relative to the known pack of boards the comparison step identified the match.