Marker Generation Device for Accurate Feature Point Matching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing marker generation technologies fail to detect identifiers as markers due to mismatched feature points, preventing accurate identification and distinction of products with different manufacturers.

Innovation Solution

A marker generation device and system that arranges feature points in a predetermined space, counts matching points, and selects identifiers with the highest count as detection markers by converting their configuration to match unique features.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If identifiers are used as markers without configuration conversion, then the original identifier structure is preserved, but the feature points do not match with unique features preventing accurate detection

Engineering Contradiction:
Improvedetection accuracyVSAvoididentifier processing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies parameter changes by converting the configuration of identifiers to match unique features. Specifically, the system extracts feature points from identifiers, compares them with unique features extracted from background images, and transforms the identifier configuration (such as scaling, rotating, or repositioning) to maximize the number of matching feature points with unique features, thereby enabling reliable detection while maintaining identifier functionality

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent implements preliminary action by pre-extracting unique features from background images before identifier detection occurs. These unique features are stored and used as reference patterns for subsequent identifier detection, allowing the system to quickly determine whether detected feature points correspond to actual markers or background elements without performing complex analysis during the detection phase

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If feature points are extracted and arranged in predetermined space to identify unique features, then accurate marker detection is enabled, but the process becomes complex when handling multiple identifier configurations

Engineering Contradiction:
Improvefeature point matching precisionVSAvoidfeature comparison system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies local quality by focusing the comparison process on specific local features rather than analyzing the entire identifier image globally. The system extracts discrete feature points from identifiers and compares their local characteristics (such as position, orientation, and distance relationships) with locally extracted unique features from background images, enabling precise matching while reducing overall computational complexity

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent implements segmentation by dividing the identifier image into discrete feature points rather than treating it as a continuous image. Each feature point is independently extracted and its coordinates are arranged in a predetermined space for comparison with unique features. This segmentation approach simplifies the complexity of handling multiple identifier configurations by reducing them to a set of comparable discrete points

Inventive Principle:
Principle #1Segmentation

3Measurement precision

If identifier configuration is converted to maximize feature point matching, then detection accuracy improves, but the original identifier information may be altered

Engineering Contradiction:
Improvedetection precisionVSAvoidoriginal identifier information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent applies copying by creating a transformed copy of the identifier configuration for detection purposes while preserving the original identifier information. The system extracts feature points from the original identifier, generates a converted configuration that maximizes matching with unique features, and uses this copy for detection without modifying the source identifier data, thus maintaining both detection accuracy and information integrity

Inventive Principle:
Principle #26Copying

Data Source

PatentEP2458558B1Marker generation device, marker generation detection system, marker generation detection device, marker, marker generation method, and program therefor
Publication Date: 2022.09.21 NEC CORP
  • EP2458558B1 patent drawingFigure 1~2
  • EP2458558B1 patent drawingFigure 3
  • EP2458558B1 patent drawingFigure 4

AI summary

Provided is a marker generation device which has a feature comparison means and a marker pattern generation means. The feature comparison means disposes feature points extracted from an image in a predetermined space, sets parts in which the number of feature points in the predetermined space is equal to or less than a predetermined number as singular features, disposes feature points extracted from an identifier in the predetermined space, and counts the number of feature points that coincide with the singular features. The marker pattern generation means converts the structure of the identifier and extracts the feature points from the converted identifier. The feature comparison means disposes the feature points extracted from the converted identifier in the predetermined space, counts the number of feature points that coincide with the singular features, and selects, as a marker for detection, an identifier having the most counts from among identifiers before conversion and one or two or more identifiers that have been converted.