Traffic Light Detection Using Alternating Exposure Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving vehicles face challenges in recognizing red traffic light signals, especially under dark or cloudy conditions due to color artifacts, leading to failures in detection and classification.
Innovation Solution
The method involves applying a first sensor setting with reduced exposure time or gain to capture a frame focused on traffic light color recognition and a second setting with normal exposure for perceiving the driving environment, allowing the vehicle to determine the traffic light color and navigate safely.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If normal exposure time is used for image capture in dark environments, then the driving environment can be perceived, but red traffic light signals cannot be recognized due to color artifacts
Solution Approach 1:
The patent segments the image capture process into two distinct modes: a first image captured with first sensor settings (reduced exposure time) optimized for traffic light color recognition, and a second image captured with second sensor settings (normal exposure time) optimized for driving environment perception. This segmentation allows each image to be optimized for its specific purpose, resolving the contradiction between color accuracy and environmental visibility.
Solution Approach 2:
The patent applies different sensor settings (different exposure times) to capture different types of information: reduced exposure time for the specific task of traffic light color recognition and normal exposure time for general driving environment perception. This local quality approach ensures that each capture mode has the optimal settings for its specific function.
2Measurement precision
If reduced exposure time is used for traffic light detection, then red traffic light signals can be recognized, but the driving environment perception is compromised
Solution Approach 1:
The patent divides the sensing task into two separate image captures: one with reduced exposure time dedicated to traffic light color recognition and another with normal exposure time dedicated to driving environment perception. This segmentation ensures that traffic light detection accuracy is not compromised while still maintaining adequate environment visibility through the second image.
Solution Approach 2:
The sensor system performs multiple functions by capturing two different types of images with different settings: one optimized for color recognition and another optimized for environmental perception. This multi-functionality allows the system to achieve both traffic light detection accuracy and driving environment visibility simultaneously.
3Adaptability or versatility
If multiple sensor settings are alternated for image capture, then both traffic light color and driving environment can be perceived, but the system complexity increases
Solution Approach 1:
The patent implements periodic alternation between first and second sensor settings for capturing images. The sensor controller switches between different exposure time settings in a regular pattern, allowing the system to periodically capture both traffic light-optimized images and environment-optimized images. This periodic action provides a systematic way to manage the complexity of multiple sensor settings while maintaining comprehensive detection capability.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A driving environment is perceived based on sensor data obtained from a plurality of sensors mounted on the ADV, including detecting a traffic light, where the plurality of sensors includes at least one image sensor. A first sensor setting (402) is applied to the at least one image sensor to capture a first frame, and a second sensor setting (403) is applied to the at least one image sensor to capture a second frame. A color of the traffic light is determined based on sensor data of the at least one image sensor in the first frame. The ADV is controlled to drive autonomously according to the color of the traffic light determined based on sensor data of the at least one image sensor in the first frame and a driving environment perceived based on sensor data of the at least one image sensor in the second frame.