Vehicle LED Flicker Mitigation via Multi-Vehicle Coordination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
LED flicker poses a challenge for Advanced Driver Assist Systems (ADAS) and Autonomous Vehicles (AV) by causing misinterpretation of traffic signal lights, leading to ambiguity in the state of LEDs, which can result in missed hazards or false readings due to the vehicle perception system's inability to accurately recognize the state of LED signals.
Innovation Solution
A vehicle-based LED flicker mitigation system that uses vehicle-to-vehicle communication to coordinate image capture timing and exposure parameters, allowing multiple vehicles to jointly observe LED light sources and correct erroneous state information by sharing image data, thereby mitigating flicker artifacts and improving the accuracy of LED state determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional LED flicker mitigation methods (LOFIC, chopping, split pixel imaging) are used, then the probability of accurately characterizing LED state increases, but system complexity and vehicle cost increase
Solution Approach 1:
The patent combines image captures from multiple vehicles to collectively characterize LED state. By merging data from multiple perception systems, the solution achieves accurate LED state detection without requiring complex hardware modifications in individual vehicles, thus resolving the contradiction between measurement precision and device complexity
Solution Approach 2:
The patent introduces communication between vehicles as an intermediary mechanism to share image capture data. This mediator approach allows vehicles to pool their observation data to accurately determine LED state without each vehicle needing complex flicker mitigation hardware, thereby reducing individual system complexity while maintaining high measurement precision
2Measurement precision
If conventional LED flicker mitigation methods are used, then LED state detection accuracy improves, but other artifacts such as non-continuous intensity (pulsing) appear in images
Solution Approach 1:
By combining image captures from multiple vehicles with different exposure timings, the system achieves continuous intensity representation. The merging of multiple perspectives eliminates the pulsing artifacts that appear in individual captures, as the combined data reconstructs the true continuous LED intensity profile
Solution Approach 2:
The patent adds a temporal dimension by utilizing image captures taken at different exposure times. This dimensional approach allows the system to observe LED state across multiple time points, thereby eliminating temporal artifacts like pulsing while maintaining accurate LED state detection
3Device complexity
If single vehicle image capture is used, then system simplicity is maintained, but LED flicker causes misinterpretation of traffic signal state
Solution Approach 1:
The patent introduces inter-vehicle communication as a mediator that enables simple individual vehicles to achieve reliable collective detection. Each vehicle maintains simplicity, but through communication and data sharing, the group achieves high reliability in LED state recognition, eliminating flicker-induced misinterpretations
Solution Approach 2:
The system uses feedback from multiple vehicle observations to correct individual perception errors. By comparing and cross-validating LED state detections across multiple vehicles, the system identifies and corrects misinterpretations caused by flicker, thereby maintaining simplicity while improving reliability
Data Source
AI summary
A light emitting diode (LED) flicker mitigation system disposed on a vehicle captures multiple images of an exterior scene to determine whether a light that appears to be off is actually off, or whether the state of the light is actually on but erroneously shown in the system image captures as off due to temporal aliasing. The LED flicker mitigation system may obtain information from external infrastructure and contact other vehicles proximate to the captured scene to determine the true state of the LED light source by identifying homography using the vehicle image captures, infrastructure information, and vehicle sensors. The images may be shared amongst the vehicles to correctly determine the LED status and correct any erroneous LED state information.