Road Curvature Estimation Using Focus Position Ahead of Hidden Lane Lines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing road curvature estimation systems face challenges in accurately estimating curvature when lane lines at distant points are obscured by obstacles like preceding or oncoming vehicles, leading to inaccurate curvature calculations.
Innovation Solution
A road curvature estimation device utilizing a stereo camera system with a main and subcamera to capture images, combined with machine learning for boundary line estimation, and a curvature calculation device to detect maximum curvature positions and set focus positions based on vehicle speed, ensuring accurate curvature estimation even when lane lines are obscured.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If curvature is estimated based on boundary lines at distant points, then curvature estimation accuracy is improved, but measurement reliability deteriorates when lane lines are obscured by obstacles
Solution Approach 1:
The system performs preliminary actions by detecting the maximum curvature position and setting a focus position ahead of the vehicle before curvature estimation is performed. This allows the system to pre-determine the optimal estimation point, ensuring that curvature is measured at a location where boundary lines are visible and reliable, even when distant lane lines are obscured by obstacles.
Solution Approach 2:
Instead of uniformly estimating curvature across all boundary line positions, the system applies local quality by identifying the maximum curvature position and setting the focus position specifically at or ahead of that location. This localized approach ensures that curvature estimation is performed at the most critical and reliable point, improving both precision and reliability.
2Measurement precision
If focus position is set at maximum curvature point, then curvature estimation precision is improved, but active lane keep control stability deteriorates due to lack of advance warning
Solution Approach 1:
The system performs preliminary action by setting the focus position ahead of the maximum curvature position when the maximum curvature point is beyond a predetermined distance. This advance positioning provides early warning to the active lane keep control system, allowing it to prepare for upcoming curvature changes and maintain stable control by anticipating rather than reactively responding to curve entries.
Solution Approach 2:
The system dynamically adjusts the focus position based on the distance to the maximum curvature position. When the maximum curvature is far ahead, the focus position is set earlier to provide advance warning. When the maximum curvature is closer, the focus position is set at the maximum curvature point itself for precise estimation. This dynamic adjustment optimizes both control stability and estimation precision under different driving conditions.
3Speed
If curvature is estimated using single frame image data, then processing speed is improved, but measurement reliability deteriorates when lane lines are hidden
Solution Approach 1:
The system performs preliminary action by detecting the maximum curvature position and setting the focus position before curvature estimation is performed. This pre-processing step allows the system to identify the optimal estimation point in advance, ensuring that single-frame curvature estimation is performed at a location where boundary lines are most likely to be visible and reliable, thereby maintaining both speed and reliability.
Solution Approach 2:
The system changes the parameter of focus position based on the detected maximum curvature position and vehicle speed. By dynamically adjusting this parameter, the system ensures that curvature estimation is performed at the most reliable location in each frame, maintaining high reliability even when processing only single-frame image data at high speed.
Data Source
AI summary
A road curvature estimation device includes a camera device and a curvature calculation device. The camera device includes an image capturer and a boundary line estimator. The curvature calculation device includes a maximum curvature position detector, a focus position setter, and a focus position curvature calculator. The focus position curvature calculator is configured to compare a maximum curvature position detected by the maximum curvature position detector and a focus position set by the focus position setter. The focus position curvature calculator is configured to, when the focus position is closer to a vehicle than is the maximum curvature position, set a curvature at the focus position as a focus position curvature. The focus position curvature calculator is configured to, when the focus position is ahead of the maximum curvature position, set a curvature at the maximum curvature position as the focus position curvature.