Lane Detection Using Coordinate-Converted Previous Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lane detection technologies face challenges in accurately estimating lanes when lane marks are faded, hidden, or temporarily undetectable due to factors like backlight or oncoming headlights, leading to degraded precision in straight line determination and inability to perform steering assistance.
Innovation Solution
A lane detection device that uses an imaging device to capture road images, detects lane marks, and employs a lane estimation unit for reliable lane estimation through curve approximation by superimposing current and previous lane data after coordinate conversion, incorporating vehicle speed and yaw rate, and a curve starting point detection unit that estimates the curve starting point using Gaussian distribution functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If lane detection is performed based on image data from an imaging device using least-square method, then lane detection can be performed in real-time, but precision degrades when lane marks are faded, hidden, or temporarily undetectable
Solution Approach 1:
The system performs preliminary actions by detecting and storing lane information in advance when lane marks are visible and detectable. This stored lane information is then used to supplement or replace current detections when lane marks become faded, hidden, or undetectable due to adverse conditions like backlight or oncoming headlights, ensuring continuous reliable lane detection.
Solution Approach 2:
The system creates a copy of lane information by storing detected lane data in memory for future use. This copied lane information from previous time points is reused when current lane mark detection fails or degrades, maintaining detection precision without requiring real-time visibility of physical lane marks.
2Device complexity
If only current lane mark detection is used for steering assistance, then system complexity is low, but steering assistance cannot be provided when lane marks are undetectable
Solution Approach 1:
The system performs preliminary detection and storage of lane information before it is needed for steering assistance. By accumulating lane data in advance and storing it in memory, the system ensures that steering assistance can continue to function even when current lane marks are temporarily undetectable, without adding significant complexity to the detection system.
3Measurement precision
If curve starting point detection is performed using traveling trace and steering angle of preceding vehicle, then accurate curve entry position can be determined, but steering assistance cannot be provided when preceding vehicle information is unavailable
Solution Approach 1:
The system uses an intermediary approach by detecting curve starting points directly from lane mark geometry and vehicle state data when preceding vehicle information is unavailable. The curve starting point detection unit identifies the point where the lane mark curvature changes significantly, serving as an intermediary method to maintain reliable steering assistance without depending solely on preceding vehicle communications.
Data Source
AI summary
A more accurate lane is estimated by using a lane at a previous time point after coordinate conversion that is coordinate-converted based on traveling information (a vehicle speed and a yaw rate) at a current time point, as well as a lane at the current time point.