Package Dimensioner Pose Change Detection via Depth Map Comparison

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing package dimensioning systems face accuracy issues due to movement of the camera or measurement platform relative to each other, which can occur unintentionally over time, leading to invalidation of the initial setup and reduced measuring accuracy.

Innovation Solution

The system employs background modeling and depth map comparison to detect significant changes in the pose of the dimensioning system by capturing initial and subsequent images, generating reference and current depth maps, and counting pixels that differ by a threshold, triggering a re-initialization process if a significant change is detected.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the camera or measurement platform moves relative to each other after initialization, then the system becomes more adaptable to environmental changes, but the measurement precision deteriorates due to invalidation of the initial setup

Engineering Contradiction:

Improvesystem adaptability to environmental changesVSAvoiddimensioning accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system performs preliminary background modeling during initialization to establish a reference state. This background model is stored and later used for comparison to detect any deviations caused by movement, allowing the system to identify when re-initialization is needed to maintain measurement precision

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system continuously compares current depth maps against the stored background model and provides feedback through alerts when significant changes are detected. This feedback mechanism enables the system to respond to environmental changes by triggering re-initialization, thus maintaining measurement accuracy while adapting to new conditions

Inventive Principle:

Principle #23Feedback

2Measurement precision

If the system continuously monitors for scene changes, then the measurement precision is maintained, but the use of energy increases due to continuous image processing

Engineering Contradiction:

Improvedimensioning accuracyVSAvoidenergy consumption for image processing

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

Instead of continuously processing all images, the system performs partial monitoring by comparing only key depth map features against the background model at intervals. This selective monitoring approach maintains measurement precision while significantly reducing the computational energy required compared to continuous full-frame processing

Inventive Principle:

Principle #16Partial or excessive action

3Reliability

If the system performs frequent scene change detection, then the reliability of measurements is improved, but the productivity decreases due to repeated re-initialization

Engineering Contradiction:

Improvemeasurement reliabilityVSAvoiddimensioning throughput

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system performs scene change detection periodically rather than continuously, comparing current depth maps against the background model at scheduled intervals. This periodic monitoring maintains measurement reliability while minimizing the frequency of re-initialization events that would reduce productivity

Inventive Principle:

Principle #19Periodic action

Data Source

PatentEP3255376B1Method for detecting a scene change in a dimensioner and corresponding dimensioning system

Publication Date: 2019.01.23 HAND HELD PRODS INC

Figure 1~6
Figure 3
Figure 4

AI summary

A package dimensioner is disclosed. A change in the pose of the package dimensioner is detected by background modeling the area of a measurement platform and then determining if a number of points in a scene are different in distance from the background model. Change in the pose can also be detected by comparing a count of support points in a 3D container generated from images taken in a training process with a count of support points in a subsequent images and determining how many support points are different.