Adaptive Lane Marker Detection via Contrast Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle vision systems face challenges in efficiently extracting lane markings, particularly in distinguishing between straight and curved lines, and in processing both black and white lines, which are computationally expensive and limited in generalization.
Innovation Solution
A unified method using gradient processing and the Hough transform is replaced by a system that determines contrasts at both sides of a detected marker over various angles, allowing for the extraction of both straight and curved lines, including black and white markings, using a harmonized approach that is computationally cheaper and more suitable for embedded applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If gradient processing followed by Hough transform is used to extract lane markings, then lane marking detection capability is improved, but computational cost increases and generalization is limited
Solution Approach 1:
The patent segments the lane marking detection process into distinct phases: gradient computation at pixel level, edge pixel identification, and line segment extraction. This segmentation allows each phase to be optimized independently, reducing overall computational complexity while maintaining detection precision.
Solution Approach 2:
The patent introduces adaptive parameters that dynamically adjust during processing, such as varying gradient thresholds and line fitting parameters based on local image characteristics. This dynamic approach enables the system to handle diverse marking types (straight, curved, black, white) without requiring multiple fixed processing pipelines.
2Adaptability or versatility
If traditional methods are used to extract lane markings, then straight lines can be detected, but curved lines and different color markings cannot be extracted without generalization
Solution Approach 1:
The patent creates a universal processing framework that handles multiple marking types (straight lines, curved lines, black markings, white markings) through a single unified algorithm. The gradient-based edge detection and adaptive line fitting methodology works universally across different marking geometries and colors, eliminating the need for separate processing pipelines for each marking type.
Solution Approach 2:
The patent employs parameter changes to adapt to different marking characteristics. By adjusting gradient thresholds, edge detection sensitivity, and line fitting parameters based on local image properties, the system can extract various marking types without changing the fundamental processing approach, thus maintaining low complexity while achieving high versatility.
3Area of stationary object
If wide angle field of view camera is used to capture images, then field of view is improved, but straight lines appear curved depending on position in the field of view
Solution Approach 1:
The patent applies dynamic line fitting that adapts to the local geometry of detected edges. Rather than assuming straight lines, the system uses adaptive curve fitting that can model both straight and curved segments, automatically adjusting to the actual marking geometry observed in the wide-angle image, thus correcting for perspective distortion.
Solution Approach 2:
The patent changes the mathematical model parameters from fixed straight-line equations to adaptive curve equations when processing wide-angle images. By allowing the line representation to transition between straight and curved models based on edge pixel distribution, the system maintains measurement precision across the entire wide field of view.
Data Source
AI summary
A method for determining lane markers includes providing a camera at a vehicle so as to have a field of view exterior of the vehicle. An image processor processes a frame of image data captured by the camera to determine intensity gradient information of captured image data and to determine lane markers. Contrast values at both sides of a center region of the determined lane marker are determined via processing of the intensity gradient information. An angle of the determined lane marker relative to the direction of travel of the vehicle is determined responsive to the determined contrast values. Processing of a subsequent frame of captured image data is adjusted responsive to the determined angle of the determined lane marker relative to the direction of travel of the vehicle.