Multi-Camera Robot Mapping Using Straight-Line Visual Features

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing localization methods for mobile robots in unstructured outdoor environments face challenges due to the limited precision of publicly available GPS systems and the dynamic nature of these environments, making it difficult for robots to accurately navigate and map their surroundings.

Innovation Solution

The use of multiple cameras on a mobile robot to take visual images, extract straight lines, and generate map data, combined with data from other sensors like GPS and Lidar, to build a precise map of the environment and localize the robot simultaneously.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If GPS receivers are used for localization, then the system can operate in outdoor environments, but the precision is limited to 1-10 meters which is insufficient for autonomous robotic navigation

Engineering Contradiction:
Improvelocalization precisionVSAvoidenvironmental adaptability
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent combines GPS receivers with visual sensors (cameras) and other sensors to create a hybrid localization system. The GPS provides coarse location information while visual features provide fine-grained localization, achieving centimeter-level precision by merging multiple data sources rather than relying on GPS alone.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

Visual features extracted from camera images serve as an intermediary between the GPS system and the robot's precise position. The visual localization system bridges the gap by using detected straight lines and visual landmarks to refine the coarse GPS coordinates into accurate robot pose information.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If visual sensors are used for mapping and localization, then the system can operate in dynamic environments with people moving, but the complexity of processing visual data increases

Engineering Contradiction:
Improveenvironmental adaptabilityVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent extracts only the most salient visual features - straight lines - from camera images for mapping and localization. By focusing on this specific feature type rather than processing all visual information, the system reduces computational complexity while maintaining effectiveness in dynamic environments.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent replaces complex mechanical laser rangefinders with visual sensors (cameras) for mapping purposes. Visual techniques provide a simpler, more economical solution that can handle dynamic environments where people move, substituting optical processing for mechanical scanning systems.

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

3Measurement precision

If multiple cameras are used to capture visual images, then the coverage and accuracy of map generation improve, but the quantity of data to be processed increases

Engineering Contradiction:
Improvemap generation accuracyVSAvoiddata volume
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent extracts only straight line features from the multiple camera images rather than processing all pixel data. This selective feature extraction dramatically reduces the data volume while preserving the geometric information needed for accurate map generation and localization.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent divides the visual processing task by having each camera independently extract straight lines from its field of view, then combining these segmented feature sets. This segmentation approach allows parallel processing of data from multiple cameras while maintaining manageable data volumes through feature-level rather than pixel-level processing.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3825807B1Method, device and assembly for map generation
Publication Date: 2025.04.02 STARSHIP TECH OU
  • EP3825807B1 patent drawingFigure 1
  • EP3825807B1 patent drawingFigure 2
  • EP3825807B1 patent drawingFigure 3

AI summary

Disclosed is a mapping method comprising operating at least one mobile robot comprising at least two cameras and at least one processing component; and taking visual images with at least two cameras; and extracting at least straight lines from the visual images with at least one processing component; and generating map data using at least the extracted straight lines from the visual images. Also disclosed is a mobile robot comprising at least two cameras adapted to take visual images of an operating area; and at least one processing component adapted to at least extract straight lines from the visual images taken by the at least two cameras and generate map data based at least partially on the images; and a communication component adapted to at least send and receive data, particularly image and/or map data. Further disclosed is an assembly of mobile robots comprising at least two mobile robots, and each of the robots comprising at least two cameras adapted to take visual images of an operating area; at least one processing component adapted to at least extract straight lines from the visual images taken by the at least two cameras and generate map data based at least partially on the images; and at least one communication component adapted to at least send and receive data at least between the robots, particularly image and/or map data.