Container Inspection Image Reconstruction Segmentation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing detection systems for containers face accuracy and efficiency issues due to the separation of image reconstruction and segmentation into independent steps, leading to increased time and cost, and the need for complex heuristic corrections to compensate for errors in reconstructed images.

Innovation Solution

A method that combines image reconstruction and segmentation into a single operation, using simulated scan data to calculate an error term and iteratively refine the reconstructed image until a satisfaction criterion is met, thereby reducing artifacts and the need for heuristic corrections.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If image reconstruction and segmentation are performed as separate independent steps, then the inspection process can be simplified in terms of algorithm design, but the accuracy of object detection decreases and inspection time increases

Engineering Contradiction:

Improvealgorithm design complexityVSAvoidobject detection accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent combines image reconstruction and segmentation into a single integrated algorithm that processes X-ray data and produces segmented images simultaneously. This merging eliminates the sequential dependency between reconstruction and segmentation steps, allowing the system to achieve both accurate object detection and reduced inspection time without requiring complex post-processing corrections.

Inventive Principle:

Principle #5Merging (Combining)

2Ease of manufacture

If image reconstruction and segmentation are performed as separate independent steps, then each step can be optimized independently, but the total inspection time and cost increase

Engineering Contradiction:

Improvealgorithm optimization flexibilityVSAvoidinspection speed

Core Design Contradiction:

Ease of manufactureVSProductivity

Solution Approach 1:

The integrated algorithm merges reconstruction and segmentation operations into a single processing pipeline that operates simultaneously on X-ray data. This eliminates the sequential time cost of performing reconstruction followed by segmentation, thereby increasing inspection speed while maintaining the ability to optimize the unified algorithm's performance.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent implements continuous processing where the algorithm continuously refines the segmentation based on reconstruction data without requiring discrete sequential steps. This continuous action eliminates idle time between steps and maintains productive processing throughout the entire inspection cycle.

Inventive Principle:

Principle #20Continuity of useful action

3Ease of operation

If known image reconstruction algorithms are used, then the reconstruction process can be optimized for human interpretation, but artifacts are introduced that require complex heuristic corrections during inspection

Engineering Contradiction:

Improvehuman interpretation optimizationVSAvoidimage artifacts

Core Design Contradiction:

Ease of operationVSObject-generated harmful factors

Solution Approach 1:

The patent converts the harmful artifacts introduced by standard reconstruction algorithms into beneficial information by incorporating them directly into the segmentation process. The integrated algorithm uses reconstruction artifacts as input features for segmentation, transforming what would be erroneous data into useful cues for identifying objects and structures within the container.

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

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

This approach enhances the accuracy and reduces the time and cost of detecting objects within containers by integrating image reconstruction and segmentation, minimizing artifacts and the requirement for complex corrections.

Implementation Method 1

Each pixel in the reconstructed image generally represents a relative X-ray attenuation caused by the contents of the container

Methodology Applied

Scientific EffectX-ray: X-Ray

Data Source

PatentUS8180138B2Method and system for inspection of containers

Publication Date: 2012.05.15 MORPHO DETECTION LLC

AI summary

A method and system for reconstructing and segmenting an image of the contents of a container. The method includes receiving actual scan data from a scan of the container, reconstructing the actual scan data to obtain a reconstructed image, and segmenting the reconstructed image to obtain a segmented image. The method also includes deriving simulated scan data corresponding to the segmented image, calculating an error term based on a difference between the simulated scan data and the actual scan data, and determining whether a criterion is satisfied. The method further includes using the error term to produce a modified reconstructed image and repeating the preceding steps with the modified reconstructed image substituted for the reconstructed image if the criterion is not satisfied, and outputting information from the segmented image if the criterion is satisfied.