Polygon Homogeneity Evaluation for Asymmetric Shape Matching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current object detection systems face challenges in accurately comparing and evaluating the homogeneity of polylines or polygons extracted from images, particularly in autonomous driving and IoT applications, where precise localization and mapping require high accuracy and real-time feature matching.

Innovation Solution

A computer vision system that employs a homogeneity metric sensitive to polyline/polygon similarity and purity, allowing for asymmetric evaluation by representing each set of polylines/polygons using the other set, enabling a finer measure of granularity in homogeneity evaluation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional object detection systems compare extracted polylines or polygons with previous detections, then object identification can be achieved, but the accuracy of homogeneity evaluation deteriorates due to lack of sensitive similarity measurement

Engineering Contradiction:
Improvehomogeneity evaluation accuracyVSAvoidfeature matching reliability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent introduces a homogeneity metric that transforms the comparison approach by changing the parameters used for evaluation. Instead of traditional similarity measures, it computes homogeneity based on pixel-level overlap and boundary alignment between detected objects and reference polylines, enabling sensitive detection of subtle differences in object representations

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent replaces traditional mechanical or geometric comparison methods with a computational homogeneity metric that uses pixel-level analysis. This substitution enables more precise measurement by leveraging image processing capabilities rather than simple coordinate-based comparison

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If polyline homogeneity is evaluated using traditional methods, then processing speed can be maintained, but measurement precision of similarity deteriorates

Engineering Contradiction:
Improvepolyline similarity measurementVSAvoidhomogeneity evaluation complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the homogeneity evaluation into distinct computational steps: rendering detected objects to polylines, comparing with reference polylines, computing pixel-level overlap, and calculating homogeneity metrics. This segmentation makes the complex evaluation process manageable and implementable in real-time systems

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary representation layer where detected objects are rendered to polylines before comparison with reference polylines. This intermediary step facilitates precise homogeneity measurement by providing a standardized format for comparison while maintaining computational efficiency

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3550471B1Method, apparatus, and system for determining polyline homogeneity
Publication Date: 2023.10.04 HERE GLOBAL BV
  • EP3550471B1 patent drawingFigure 1
  • EP3550471B1 patent drawingFigure 2
  • EP3550471B1 patent drawingFigure 3

AI summary

An approach is provided for an asymmetric evaluation of polygon similarity. The approach, for instance, involves receiving a first polygon representing an object depicted in an image. The approach also involves generating a transformation of the image comprising image elements whose values are based on a respective distance that each image element is from a nearest image element located on a first boundary of the first polygon. The approach further involves determining a subset of the plurality of image elements of the transformation that intersect with a second boundary of a second polygon. The approach further involves calculating a polygon similarity of the second polygon with respect the first polygon based on the values of the subset of image elements normalized to a length of the second boundary of the second polygon.