ULD Type Identification via Template Matching and Grid Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current imaging systems in commercial shipping struggle to accurately and efficiently determine the type of unit load devices (ULDs) at load points, often leading to incorrect or missing analytics due to manual errors in barcode scanning and load point ID entry.
Innovation Solution
A method and system that capture image data of ULDs, align it with templates, convert to down-sampled grids, remove non-dense areas, and calculate match scores to identify the ULD type by determining the shortest distance between grid values in the ULD and template borders, using techniques like k-d tree search and depth-first search algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual barcode scanning and load point ID entry are used, then the system is simple to operate, but accuracy and reliability of ULD container type determination deteriorate due to human error
Solution Approach 1:
The patent replaces the manual mechanical process of barcode scanning and ID entry with an automated imaging system that captures images of the ULD and uses image processing algorithms to automatically determine container type. This substitution eliminates human error while maintaining operational simplicity through automation.
Solution Approach 2:
The system creates a digital copy (image) of the physical ULD container and processes this copy through template matching algorithms to identify the container type. This copying approach allows for accurate, repeatable measurements without requiring manual intervention.
2Productivity
If traditional imaging systems are used, then device complexity is low, but productivity and speed of ULD container type assessment deteriorate
Solution Approach 1:
The system performs preliminary actions by pre-processing the captured image data through alignment and down-sampling before the actual container type determination. This preliminary processing prepares the data for faster and more accurate template matching, improving overall productivity.
Solution Approach 2:
The image processing is segmented into distinct stages: image capture, alignment, down-sampling, border identification, and template matching. This segmentation allows each stage to be optimized independently, improving processing speed while managing system complexity.
3Loss of time
If manual scanning processes are used, then ease of operation is high, but loss of time due to human error and rework increases
Solution Approach 1:
The system performs self-service by automatically capturing images, processing them through template matching, and determining container types without requiring operator intervention. This eliminates time losses associated with manual scanning errors and rework while maintaining ease of operation through automated workflows.
4Measurement precision
If high-fidelity container analytics are implemented, then measurement precision improves, but device complexity increases due to multiple ULD types requiring different algorithms
Solution Approach 1:
The template matching system provides a universal solution that can determine container types across multiple ULD types using a single algorithmic approach. The system maintains high measurement precision by using comprehensive templates that cover various container types, eliminating the need for separate algorithms for each ULD type.
Data Source
AI summary
Methods for determining a unit load device (ULD) container type are disclosed herein. An example method includes capturing a set of image data featuring the ULD and aligning the set of image data with a template. The method further includes converting the set of image data and the template to down-sampled grids including a plurality of rows and columns. The method further includes removing portions of the image data grid that do not exceed a density threshold. The method further includes identifying a ULD border and a template border by extracting leftmost, rightmost, and topmost grid values from the respective grids. The method further includes calculating a match score corresponding to the template by determining a shortest respective distance between grid values in the ULD border and the template border, and determining ULD container type corresponding to the ULD based on the match score.