Vehicle Lane Estimation Using Integrated Detector Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional traveling lane estimation methods suffer from reduced reliability due to decreased sensor accuracy in certain environments, and using multiple detection methods with varying accuracy can adversely affect the final estimation result.
Innovation Solution
A vehicle control device integrates the reliability calculations of multiple detectors, including a first, second, third, and fourth detector, to estimate a traveling lane by comparing their integrated reliabilities, ensuring accurate estimation even if individual detectors have insufficient accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If autonomous driving control is implemented without driver input detection, then automation level is improved, but safety is worsened due to inability to detect driver takeover intent
Solution Approach 1:
The system continuously monitors driver state through imaging devices and detects takeover intent through facial expression analysis, providing feedback to the autonomous driving control system to switch between autonomous and manual modes appropriately
Solution Approach 2:
An intermediary system comprising imaging devices and analysis units is introduced between the autonomous driving control and the driver, enabling indirect detection of driver takeover intent through facial expressions without requiring direct driver input
2Reliability
If driver input detection is added to autonomous driving system, then safety is improved, but device complexity is worsened
Solution Approach 1:
The imaging device serves multiple functions: monitoring driver state, detecting takeover intent through facial expressions, and providing feedback to the control system, eliminating the need for separate detection devices
Solution Approach 2:
The system uses the driver's own facial expressions as the detection source, requiring no additional sensors or input devices from the driver, thereby simplifying the overall system architecture
3Measurement precision
If facial expression analysis is used to detect takeover intent, then detection precision is improved, but loss of time is worsened due to analysis processing time
Solution Approach 1:
The system continuously captures and pre-processes facial expression data in real-time, so that when takeover intent is needed, the analysis can be performed immediately on already-processed data, reducing detection latency
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided is a technique that when a traveling lane is determined by using a plurality of detection methods, can accurately estimate the traveling lane even if any of the detection methods is not sufficient in accuracy. A vehicle control device according to the present invention calculates integrated reliabilities by integrating reliabilities of lane estimation results by a first detector and reliabilities of lane estimation results by a second detector, and estimates a lane in which a vehicle is traveling by comparing the integrated reliabilities with each other.