Road Marking Intersection Detection With Search Region Restriction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting intersection points of road marking lines require high computational complexity, which is not suitable for real-time applications in autonomous or partially autonomous driving due to limited computation resources in embedded systems.
Innovation Solution
A multistage detection method is employed, where a road marking line is first identified, and a search region is automatically determined around it, followed by a restricted search for intersection points using line and feature recognition algorithms, reducing computational complexity by avoiding geometric calculations and false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a comprehensive search method is used to detect all intersection points in the entire image, then detection completeness is improved, but computational complexity increases significantly
Solution Approach 1:
The patent divides the image processing task into two stages: first detecting road marking lines, then searching for intersection points only in regions surrounding these detected lines. This segmentation reduces the search space from the entire image to specific regions of interest, thereby maintaining detection completeness while significantly reducing computational complexity
Solution Approach 2:
The patent performs preliminary detection of road marking lines before searching for intersection points. By pre-identifying the locations and orientations of road marking lines, the system can limit subsequent intersection point searches to regions surrounding these lines, avoiding unnecessary computations in areas where no intersections are likely to occur
2Measurement precision
If geometric calculations are performed to precisely determine intersection points, then measurement precision is improved, but computational time increases
Solution Approach 1:
The patent employs computationally inexpensive image processing techniques such as edge detection and line detection algorithms that provide sufficient precision for intersection point detection without requiring complex geometric calculations. These simpler methods consume fewer computational resources and execute faster, trading minor precision losses for significant gains in processing speed
3Reliability
If the search region is expanded to cover the entire image, then detection reliability is improved, but processing speed decreases
Solution Approach 1:
The patent applies different processing strategies to different regions of the image: full-image line detection is performed once to identify road marking lines, but subsequent intersection point searches are confined to local regions surrounding these lines. This localized approach maintains reliability by ensuring all potential intersections are captured while improving speed by avoiding redundant searches in irrelevant areas
Data Source
AI summary
A method for detecting at least one intersection point of road marking lines, wherein a computing unit of a transportation vehicle receives a camera image and identifies a first road marking line based on the camera image. Using the computing unit, a search region is automatically defined in an environment of the first road marking line, and an automatic search limited to the search region is performed. At least one intersection point of the first road marking line with at least a second road marking line is identified as the result of the search.