Lane And Road Boundary Detection With Multi-Stage ROI Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems for autonomous vehicles to detect lanes and road boundaries are either too computationally expensive to operate in real-time or sacrifice accuracy, and they struggle in less than ideal weather and road conditions.
Innovation Solution
A deep neural network trained with low-resolution images, region of interest images, and ground truth masks is used for real-time lane and road boundary detection, employing post-processing techniques to enhance accuracy and reduce computational requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional computer vision techniques are used for lane and road boundary detection, then measurement precision is improved, but productivity deteriorates due to high computational cost preventing real-time operation
Solution Approach 1:
The patent divides the image processing task into multiple stages: initial low-resolution processing to identify potential lane regions, followed by focused high-resolution analysis only on those specific regions. This segmentation approach maintains detection accuracy while reducing overall computational burden to enable real-time processing.
Solution Approach 2:
The system applies full computational resources selectively only to regions of interest rather than processing the entire image at high resolution. By performing partial action on relevant areas and simplified action on other areas, the system achieves real-time processing without sacrificing detection precision.
2Productivity
If computational shortcuts are implemented to enable real-time operation, then productivity is improved, but measurement precision deteriorates due to reduced accuracy
Solution Approach 1:
The patent segments the detection process into coarse initial filtering and fine detailed analysis. This allows real-time operation through efficient initial processing while maintaining accuracy through subsequent focused refinement on identified regions of interest.
Solution Approach 2:
The system processes images at multiple resolution dimensions, starting with low-resolution overview and progressing to high-resolution detailed analysis only where needed. This multi-dimensional approach enables real-time processing while preserving detection accuracy through selective high-resolution processing.
3Adaptability or versatility
If conventional detection systems are used, then they can operate in various conditions, but reliability deteriorates in less than ideal weather and road conditions
Solution Approach 1:
The system performs preliminary processing on low-resolution images to identify potential lane regions before conducting detailed analysis. This preliminary action prepares the system to focus computational resources on relevant areas, improving reliability in adverse conditions where features may be less distinct.
Solution Approach 2:
The multi-stage processing system uses feedback from the initial low-resolution analysis to guide subsequent high-resolution processing. This feedback mechanism allows the system to adapt its processing focus based on detected features, improving reliability in varying weather and road conditions.
Data Source
AI summary
In various examples, sensor data representative of an image of a field of view of a vehicle sensor may be received and the sensor data may be applied to a machine learning model. The machine learning model may compute a segmentation mask representative of portions of the image corresponding to lane markings of the driving surface of the vehicle. Analysis of the segmentation mask may be performed to determine lane marking types, and lane boundaries may be generated by performing curve fitting on the lane markings corresponding to each of the lane marking types. The data representative of the lane boundaries may then be sent to a component of the vehicle for use in navigating the vehicle through the driving surface.