3D Traffic Signal Mapping for Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robot vehicles face challenges in perceiving and interpreting traffic signals, particularly due to the need for visual detection methods that are affected by varying outdoor conditions, and existing solutions require significant infrastructure investment for broadcasting traffic light states.
Innovation Solution
A method and device for determining three-dimensional locations of traffic signals using images associated with geographic location and orientation information, identifying red, yellow, and green objects, and generating maps to facilitate real-time detection and status determination of traffic signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traffic light states are broadcast over radio, then real-time detection accuracy is improved, but infrastructure investment cost increases significantly
Solution Approach 1:
The system uses the robot vehicle's own camera and processing capabilities to detect and interpret traffic signals, eliminating the need for external radio broadcasting infrastructure. The vehicle serves itself by capturing images with its camera, processing them to identify traffic light states, and using this information for navigation decisions.
Solution Approach 2:
The patent replaces the radio broadcasting system with a vision-based detection system using cameras and image processing algorithms. Instead of receiving electronic broadcasts, the system captures optical information from traffic lights and extracts signal states through computer vision techniques.
2Quantity of substance
If vision-based traffic signal detection is used, then infrastructure investment is reduced, but detection reliability deteriorates due to varying outdoor conditions
Solution Approach 1:
The system pre-processes images by identifying and isolating traffic light regions before full analysis. It uses color segmentation to pre-identify potential traffic light locations and states, then focuses detailed processing only on these regions, improving reliability by ensuring consistent detection across varying conditions.
Solution Approach 2:
The system exploits the distinct color characteristics of traffic lights (red, yellow, green) as primary identification features. By detecting color patterns and transitions in the captured images, the system can reliably determine traffic signal states even when other environmental conditions vary.
3Measurement precision
If multiple images from different geographic locations are processed, then three-dimensional traffic signal mapping accuracy is improved, but computational complexity increases
Solution Approach 1:
The system extracts only the essential geometric and color information from multiple images needed for 3D localization, rather than processing complete image data. It identifies traffic light positions in each image, extracts color states, and uses only these key features for triangulation and mapping, reducing computational complexity while maintaining accuracy.
Solution Approach 2:
The system uses geographic location information as an intermediary to associate detected traffic lights across multiple images. By matching detected objects with their geographic coordinates and using these as intermediaries, it simplifies the correlation process between images from different positions and reduces overall computational burden.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A system and method provides maps identifying the 3D location of traffic lights. The position, location, and orientation of a traffic light may be automatically extrapolated from two or more images. The maps may then be used to assist robotic vehicles or human drivers to identify the location and status of a traffic signal (540).