Warehouse Vehicle Localization Using Ceiling Skylight Centerlines

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Industrial vehicles navigating through warehouses face challenges in localization and navigation due to the complexity of ceiling light patterns, particularly with non-rectangular skylights, which existing methods struggle to accurately interpret for precise positioning and path determination.

Innovation Solution

The implementation of a camera-mounted system on industrial vehicles that captures images of ceiling lights, processes them using convex hull determination and Hough transforms to extract centerlines, allowing for accurate vehicle positioning and navigation through warehouses with circular and merged skylights.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If traditional localization methods are used with rectangular skylights, then the system works for simple ceiling patterns, but it fails to accurately interpret complex ceiling light patterns including non-rectangular and merged skylights

Engineering Contradiction:
Improveadaptability to different ceiling light patternsVSAvoidlocalization accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent transforms the ceiling light detection problem from shape-based recognition to parameter-based analysis by extracting geometric features (convex hull, line segments, angles) and using Hough transform to convert image space features to parameter space representations. This allows the system to handle diverse ceiling light patterns including non-rectangular and merged skylights by analyzing their parametric properties rather than requiring exact shape matching

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent applies Hough transform to convert 2D image space coordinates into a different parameter space representation. By transforming the problem from direct image coordinate analysis to parameter space analysis (using line equations and geometric parameters), the system gains the ability to detect and differentiate ceiling light patterns based on their parametric characteristics rather than their raw pixel positions

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

2Measurement precision

If advanced image processing techniques are implemented to handle complex skylight patterns, then localization accuracy improves, but computational complexity and processing time increase

Engineering Contradiction:
Improvelocalization accuracyVSAvoidimage processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the complex ceiling light pattern recognition task into distinct processing stages: feature extraction (identifying individual ceiling light elements), convex hull determination (calculating geometric boundaries), line segment extraction (identifying edge characteristics), Hough transform application (converting to parameter space), and centerline determination (calculating central axes). This segmentation allows each sub-task to be optimized independently and processed efficiently

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary geometric processing by determining convex hulls and extracting line segments before applying the computationally intensive Hough transform. By pre-processing the image data to identify and simplify geometric features, the system reduces the complexity of the subsequent parameter space transformation and improves overall processing efficiency

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3489894B1Industrial vehicles with overhead light based localization
Publication Date: 2020.07.01 CROWN EQUIP CORP
  • EP3489894B1 patent drawingFigure 1
  • EP3489894B1 patent drawingFigure 2
  • EP3489894B1 patent drawingFigure 3

AI summary

According to the embodiments described herein, a method for environmental based localization may include capturing an input image of a ceiling comprising a plurality of skylights and a substantially circular light. Features can be extracted from the input image. A convex hull of the raw features can be determined, and a preferred set of lines from the raw features can be selected utilizing the convex hull. A centerline of the skylight can be determined from the preferred set of lines. The substantially circular light of the input image can be transformed into a point feature. A pose and/or position of a vehicle can be determined based upon the centerline of the skylight and the point feature.