Dual-View Traffic Light Detection for Autonomous Intersection Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in reliably detecting traffic lights, especially when the front view traffic light detection system malfunctions or is unavailable, which can lead to unintended actions like running a red light.
Innovation Solution
The implementation of a dual-view traffic light detection system, where the vehicle uses both front view cameras and independent side view sensors (including cameras, LiDAR, and Radar) to cross-check traffic light states, ensuring accurate and reliable detection even if one system fails.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single front view traffic light detection system is used, then the device complexity is low, but the reliability of traffic light detection deteriorates when the system malfunctions or is unavailable
Solution Approach 1:
The traffic light detection system is segmented into multiple independent detection modules: front view TLD system, side view TLD system, and cross traffic detection system. Each module operates independently to detect traffic light states from different perspectives, ensuring that if one module fails, others can still provide detection results.
Solution Approach 2:
The system implements beforehand cushioning by preparing backup detection pathways in advance. When the front view TLD system is unavailable, the side view TLD system is pre-configured to take over detection tasks, preventing system failure and maintaining continuous operational reliability.
2Reliability
If multiple independent detection systems are used to cross-check traffic light states, then the reliability improves, but the device complexity increases
Solution Approach 1:
Multiple independent detection systems (front view TLD, side view TLD, cross traffic detection) are merged into a unified detection framework. The systems share common processing modules including the traffic light state determination module and the vehicle control module, reducing overall system complexity while maintaining redundancy for improved reliability.
Solution Approach 2:
The detection systems are designed with universal functionality where the side view TLD system and cross traffic detection system can serve multiple purposes: primary detection when front view is unavailable, cross-checking for accuracy verification, and providing backup detection capabilities. This multi-functionality reduces the need for dedicated separate systems.
3Measurement precision
If front view TLD system result is used directly without cross-checking, then the processing speed is high, but the accuracy deteriorates when the detection is incorrect
Solution Approach 1:
The system implements partial cross-checking rather than complete verification of all detection results. When the front view TLD system detects a traffic light state, only critical discrepancies are cross-checked using side view and cross traffic detection systems, balancing accuracy improvement with time loss minimization.
Solution Approach 2:
A feedback mechanism is implemented where detection results from front view, side view, and cross traffic systems are continuously compared and cross-checked. Discrepancies trigger re-detection or alternative detection pathway activation, improving accuracy while the feedback loop learns from past detections to reduce unnecessary re-checks over time.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the reliability of traffic light detection by providing comprehensive and consistent information, preventing unintended vehicle operations and improving driving safety by ensuring accurate decision-making.
Implementation Method 1
independent side view TLD system including sensors (e.g., left/right side cameras, left/right LiDAR sensors, and/or left/right Radar sensors)
Implementation Method 2
independent side view TLD system including sensors (e.g., left/right side cameras, left/right LiDAR sensors, and/or left/right Radar sensors)
Data Source
AI summary
Provided are methods for managing traffic light detections, which can include: deriving a first state of a traffic light at an intersection a vehicle is approaching, according to first detection data acquired by a first traffic light detection (TLD) system; deriving a second state of the traffic light at the intersection, according to second detection data acquired by a second TLD system that is independent from the first TLD system; determining traffic light information at the intersection based on at least one of (i) the first state or (ii) a result of checking whether the first state is same as the second state; and causing the vehicle to operate in accordance with the determined traffic light information at the intersection. Systems and computer program products are also provided.