Autonomous Vehicle Mapping with Scan-Line Lane Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing map data for autonomous vehicles lacks accuracy and robustness in representing drivable and non-drivable surfaces, lane markings, and lane boundaries, which affects efficient vehicle navigation and maneuvering.
Innovation Solution
A system that generates map data by processing sensor data using machine learning models to detect drivable surfaces, determine polygons, medial axes, and scan lines, and cluster intersection points to accurately represent lane markings, enabling precise lane boundary detection and representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional map data collection methods are used, then map data can be generated, but the accuracy and robustness of lane marking representations is insufficient
Solution Approach 1:
The patent segments the lane marking detection process into multiple independent components: scan line generation, intersection point detection, clustering algorithms, and representation generation. Each component processes specific aspects of lane markings separately, then integrates results to create comprehensive and accurate map data with improved reliability
Solution Approach 2:
The patent employs excessive sampling by generating multiple scan lines at different orientations and positions across the drivable surface. This over-sampling ensures that lane markings are detected from multiple perspectives, and clustering algorithms process more intersection points than strictly necessary to guarantee accurate lane marking representations even under varying conditions
2Measurement precision
If detailed sensor data processing is performed to improve map accuracy, then lane marking representation quality increases, but computational complexity increases
Solution Approach 1:
The complex processing task is divided into sequential modules: polygon determination from sensor data, scan line generation, intersection point calculation, clustering, and representation generation. Each module handles a specific computational aspect, making the overall complex process more manageable and efficient while maintaining high accuracy
Solution Approach 2:
The patent performs preliminary processing by determining the drivable surface polygon and generating scan lines before actual lane marking detection. This pre-processing organizes the data structure and defines the search space, reducing the computational burden of subsequent intersection detection and clustering operations
Data Source
AI summary
Techniques for generating (e.g., creating and/or updating) map data associated with an environment based on sensor data. An example system may determine (e.g., based on the sensor data) and/or receive an indication of a lane boundary associated with a path. The system may determine, based on a set of drivable areas associated with the path, a polygon. The system may determine a representative line associated with the polygon and determine a set of points along the representative line. The system may determine a set of intersections of a set of lines that originate at the set of points and extend laterally from the representative line. The system may determine, based on the set of intersections, a representation of the lane boundary and associate the representation with the map data.