Lane Recognition Using Complementary Points and Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lane-dividing line recognition systems for vehicles face challenges in accurately generating complementary data for lane-dividing lines behind the vehicle, due to detection errors in yaw rate and other factors.
Innovation Solution
A lane-dividing line recognition apparatus that includes a processor to recognize lane-dividing lines ahead, calculate relative and absolute coordinates of tentative complementary points, and integrate first and second lane-dividing line complementary points using evaluation values for accuracy and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If complementary data is generated based on detected lane-dividing line points and vehicle traveling information, then lane recognition coverage is improved, but accuracy deteriorates due to detection errors in yaw rate and other factors
Solution Approach 1:
The patent divides the lane recognition task into two independent parts: front lane detection using camera sensors and back lane complementation using calculated complementary points. This segmentation allows each part to be optimized independently, with the front detection maintaining high accuracy while the back complementation extends coverage area.
Solution Approach 2:
The patent introduces an intermediary calculation process that uses detected lane points and vehicle traveling information (position, orientation, speed) as intermediate data to generate complementary lane points. This intermediary approach bridges the gap between detected front lane data and required back lane data, enabling accurate complementation despite detection errors.
2Measurement precision
If complementary data is generated using detected lane points and vehicle traveling information, then lane recognition accuracy is improved, but system complexity increases
Solution Approach 1:
The patent makes the lane recognition system multi-functional by enabling a single detected lane point to serve dual purposes: direct use for front lane recognition and transformation into complementary points for back lane recognition. This universality reduces the need for separate detection systems while maintaining high accuracy.
Solution Approach 2:
The patent transforms lane recognition from a static detection problem to a dynamic parameter transformation problem. By changing the reference frame and applying transformation parameters based on vehicle traveling information, the system converts front lane detections into accurate back lane complementary data without adding complex hardware.
Data Source
AI summary
A lane-dividing line recognition apparatus for a vehicle includes a processor that calculates relative coordinates of a tentative complementary point extracted from data on a lane-dividing line recognized in each predetermined cycle, based on image of a traveling environment in front of the vehicle, calculates a first lane-dividing line complementary point, based on the relative coordinates and traveling information on the vehicle, and calculates a second lane-dividing line complementary point, based on the relative coordinates and absolute position data on the vehicle. The processor further acquires a first evaluation value usable for evaluating accuracy in detecting the traveling information on the vehicle, and acquires a second evaluation value usable for evaluating accuracy in detecting the absolute position data on the vehicle, and integrates the first lane-dividing line complementary point and the second lane-dividing line complementary point, using reliability calculated based on the first evaluation value and the second evaluation value.