Lane Marking Detection Using Candidate Region Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating due to the vast amount of data they need to process and store, including visual information, GPS data, and sensor data, which can limit their navigation capabilities.
Innovation Solution
A navigation system that includes a processor and memory with instructions to analyze images from a camera onboard the vehicle using a trained model to identify candidate regions corresponding to objects, determine scores indicating the presence of lane markings, and select the appropriate candidate region.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional mapping technology is used for navigation, then the vehicle can navigate using stored map data, but the sheer volume of data needed to store and update the map poses daunting challenges
Solution Approach 1:
The patent extracts only the essential navigational elements (lane markings) from the complete map data. Instead of storing and processing entire map datasets, the system identifies and processes only lane marking information from captured images, significantly reducing data storage requirements while maintaining navigation functionality
Solution Approach 2:
The patent segments the navigation problem into identifying specific lane marking regions within images. By dividing the image into candidate regions and evaluating each region independently using trained models, the system processes only relevant information rather than analyzing complete map datasets
2Measurement precision
If vast volumes of information are collected and analyzed, then the vehicle can make accurate navigation decisions, but the sheer quantity of data limits or adversely affects autonomous navigation
Solution Approach 1:
The system extracts only essential navigational information (lane markings) from captured images rather than processing all available sensor data. This selective extraction maintains navigation accuracy by focusing on critical elements while improving processing speed and reducing computational burden
Solution Approach 2:
The patent applies preliminary action by using trained models to pre-identify candidate regions that contain lane markings before detailed analysis. This preliminary segmentation allows the system to focus computational resources only on relevant regions, improving both accuracy and processing efficiency
Data Source
AI summary
Systems and methods for identifying lane marks in an environment of a host vehicle are disclosed. In one implementation, a system includes a processor configured to receive an image acquired by a camera onboard the host vehicle and analyze the image using a trained model. The trained model is configured to identify a plurality of candidate regions, each of the plurality of candidate regions corresponding to a representation of an object in the image; determine a plurality of scores, each of the plurality of scores indicating a degree to which a corresponding one of the plurality of candidate regions corresponds to a representation of a lane marking in the image; analyze the plurality of scores to select one of the plurality of candidate regions; and output an indicator of the selected candidate region.