Code-Object Association via 3D Ray Back-Projection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing machine vision systems for reading codes on objects have limitations, such as inability to read two-dimensional symbols and limited depth of focus, which complicates the association of codes with objects, especially in conveyor systems where objects of varying heights and poses are present.

Innovation Solution

A machine vision system utilizing an area-scan camera and a dimensioner to capture images and determine object dimensions and pose, allowing for accurate association of codes with objects by transforming image coordinates into a shared coordinate space for precise surface identification and code-object correlation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a laser-scanner is used to read codes on objects, then the system can read one-dimensional barcodes, but it cannot read two-dimensional symbols

Engineering Contradiction:
Improvecode reading capabilityVSAvoidcode reading reliability
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent employs an area scan camera that can read both one-dimensional barcodes and two-dimensional symbols, replacing the specialized laser-scanner. This universal device handles multiple code types without requiring system changes, resolving the contradiction between versatility and reliability.

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Measurement precision

If a line-scan camera is used to read codes, then the system can capture code images, but it has limited depth of focus requiring additional complexity to adjust focus for objects of different heights

Engineering Contradiction:
Improvecode image capture accuracyVSAvoidfocus adjustment complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent transitions from a line-scan camera (one-dimensional scanning) to an area scan camera (two-dimensional capture). This dimensional change provides a broader depth of focus that accommodates objects of varying heights without requiring complex focus adjustment mechanisms, thereby reducing device complexity while maintaining measurement precision.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Adaptability or versatility

If an area-scan camera is used to read codes on objects of varying heights, then the system can capture images of objects at different depths, but it requires complex processing to associate codes with specific objects

Engineering Contradiction:
Improvehandling objects of varying heightsVSAvoidcode-object association processing
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent introduces a conveyor belt position system as an intermediary element. By tracking the position of objects on the conveyor belt and correlating this positional data with captured code images, the system accurately associates codes with specific objects without requiring complex image processing or object recognition algorithms, thus reducing overall system complexity.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP2833323B1Associating a code with an object
Publication Date: 2020.11.11 COGNEX CORP
  • EP2833323B1 patent drawingFigure 1A
  • EP2833323B1 patent drawingFigure 1B
  • EP2833323B1 patent drawingFigure 1C

AI summary

Described are machine vision systems, methods, and apparatus, including computer program products for associating codes with objects. In an embodiment, a machine vision system includes an area-scan camera having a field of view (FOV), the area-scan camera disposed relative to a first workspace such that the FOV covers at least a portion of the first workspace and a dimensioner disposed relative to a second workspace. The machine vision system includes a machine vision processor configured to: determine an image location of a code in an image; determine a ray in a shared coordinate space that is a back-projection of the image location of the code; determine one or more surfaces of one or more objects based on dimensioning data; determine a first surface of the one or more surfaces that intersects the 3D ray; and associate the code with an object associated with the first surface.