Traffic Light Recognition via Map-Guided Region of Interest
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current self-driving and assisted driving systems face challenges in accurately recognizing traffic lights, leading to safety concerns due to poor recognition accuracy and limited recognition distance.
Innovation Solution
A traffic light recognition system that combines positioning information from a map with real-time road images using a localization module and image processing module, generating a region of interest to enhance recognition accuracy and distance, employing computer vision and machine learning algorithms to identify traffic light status.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image recognition methods are used to identify traffic lights, then the system can process full road images, but the recognition accuracy and distance are limited
Solution Approach 1:
The patent divides the full road image into multiple sub-images based on map positioning information, creating regions of interest that contain traffic lights. This segmentation allows the system to focus computational resources on specific areas where traffic lights are likely to be found, improving recognition accuracy while reducing overall processing complexity.
Solution Approach 2:
The system performs preliminary actions by generating regions of interest before actual traffic light recognition. Using map data and positioning information, the system pre-identifies potential traffic light locations and creates focused search areas, which then guides the image recognition process to achieve better accuracy more efficiently.
2Reliability
If the system processes entire road images to ensure comprehensive coverage, then no traffic light is missed, but computational resources and processing time increase
Solution Approach 1:
The patent applies local quality by assigning different processing priorities to different regions of the road image. Regions identified as containing traffic lights based on map positioning receive focused, high-quality processing, while other areas receive minimal or no processing. This maintains detection reliability for critical elements while improving overall processing efficiency.
3Length of stationary object
If the system increases recognition distance to detect traffic lights farther away, then more advance notice is provided, but recognition accuracy decreases
Solution Approach 1:
The patent introduces a spatial dimension from map data to guide image processing. By combining 2D image data with 3D positioning information and map coordinates, the system creates vertically-oriented regions of interest that extend to expected traffic light locations. This dimensional integration enables the system to maintain high recognition accuracy at greater distances by focusing computational resources on predicted traffic light positions rather than processing entire images.
Data Source
AI summary
A traffic light recognition system including a map, a localization module, at least one image capturing device and an image processing module is provided. The map is configured to provide an information relevant to a traffic light. The localization module is configured to provide a positioning information relevant to the traffic light. At least one image capturing device is configured to capture a real-time road image relevant to the traffic light. The image processing module is configured to combine the map and the positioning information of the traffic light provided by the localization module to generate a region of interest in the real-time road image captured by the image capturing device, and to recognize the traffic light in the region of interest, wherein the traffic light includes a light box and at least one light signal.