Coordinate Registration for Contraband Detection Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Contraband detection systems face challenges in accurately and efficiently identifying regions of interest within items across different coordinate systems, leading to difficulties in precise targeting for further inspection in contraband detection systems, particularly in transitioning items from initial volumetric scanning to secondary level scanning.
Innovation Solution
The method involves forming a volumetric image in a first coordinate system, identifying regions of interest, and registering it with a projection image in a second coordinate system using image registration techniques, such as transformation computation and feature extraction, to accurately locate suspicious regions for further inspection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If volumetric imaging is used to form three-dimensional images of items, then the ability to identify regions of interest is improved, but the difficulty of coordinating between different scanning systems increases
Solution Approach 1:
The patent introduces a coordinate transformation module as an intermediary between the volumetric imaging system (first coordinate system) and the projection imaging system (second coordinate system). This intermediary performs mathematical coordinate transformations to map regions of interest from the volumetric image coordinate system to the projection image coordinate system, enabling accurate coordination without direct complex integration between the two scanning systems.
2Measurement precision
If multi-coordinate system registration is implemented to improve targeting accuracy, then the precision of suspicious region identification is improved, but the processing time increases
Solution Approach 1:
The patent performs coordinate system registration and transformation parameter calculation in advance, before the actual contraband detection process. The transformation parameters between coordinate systems are pre-computed and stored, so that during actual operation, only simple parameter application is needed rather than full re-registration, significantly reducing processing time while maintaining accuracy.
3Measurement precision
If transformation computation is used to register images from different coordinate systems, then the accuracy of region mapping is improved, but the computational complexity increases
Solution Approach 1:
The patent transforms the complex image registration problem into a parameter transformation problem. Instead of performing complex image-to-image registration computations, the system computes transformation parameters (translation, rotation, scaling) that describe the relationship between coordinate systems, then applies these parameters to map regions of interest. This parameter-based approach simplifies the computational complexity while maintaining mapping accuracy.
Data Source
AI summary
A multi-level contraband detection system. At a first level, the system obtains volumetric information about an item under inspection. The volumetric information provides a basis for identifying suspect objects in the item under inspection and their locations. The location information is expressed in a first coordinate system relative to the device used for first level scanning. When first level scanning identifies a suspicious object, the item under inspection is passed to a second level scanner that can take further measurements on the suspicious objects. The second level machine is controlled in a second coordinate system. A translation between the two coordinate systems is obtained by registering a multi-dimensional image obtained at the first level with positioning information obtained at the second level. Registration is performed using a coarse and then a fine registration process for quick and accurate registration.