Vector Map Generation From Bird's-Eye Images for Autonomous Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating high definition (HD) vector maps are computationally intensive and manual labor-intensive, and they lack coverage for all locations, with maps becoming outdated due to changes such as new road construction or maintenance, which hampers autonomous vehicle navigation.
Innovation Solution
A system that processes birds-eye view images to generate a spatial graph representation of a geographic area, using neural networks to classify pixels and extract features, transforming them into a vector map dataset, which can be used by autonomous vehicles to generate trajectories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If current methods are used to generate vector maps, then map data can be produced, but the process is computationally intensive and manual labor intensive
Solution Approach 1:
The patent replaces manual mechanical processes with an automated neural network system. The neural network automatically processes images to extract map features, eliminating the need for manual feature extraction and reducing computational complexity through automated pattern recognition.
Solution Approach 2:
The system performs self-updating by automatically detecting changes in the environment through image processing and updating the vector map without external intervention. The neural network continuously processes new images and autonomously updates map data, making the system self-maintaining.
2Adaptability or versatility
If comprehensive vector maps are created for all locations, then navigation coverage is improved, but the maps become outdated due to changes like new road construction
Solution Approach 1:
The patent implements a dynamic update mechanism where the vector map is continuously updated as new images are processed. The system detects changes in the environment (new roads, construction, etc.) and automatically updates the map data, ensuring the map remains current while maintaining comprehensive coverage.
Solution Approach 2:
The system uses feedback from continuous image processing to detect environmental changes. By comparing new images with existing map data, the neural network identifies changes and triggers automatic map updates, ensuring the map remains accurate and current.
3Ease of operation
If multiple networks and pre-processing are used to transfer data, then data can be transferred between systems, but the process becomes manual labor intensive
Solution Approach 1:
The patent creates a universal image processing pipeline that can handle multiple data sources and transfer formats. The neural network system serves multiple functions: image processing, feature extraction, and map updating, eliminating the need for separate manual processing steps for different networks.
Data Source
AI summary
A system will generate a vector map of a geographic area using a method that includes receiving a birds-eye view image of a geographic area. The birds-eye view image comprises various pixels. The system will process the birds-eye view image to generate a spatial graph representation of the geographic area, and it will save the node pixels and the lines to a vector map data set. The processor may be a component of a vehicle such as an autonomous vehicle. If so, the system may use the vector map data set to generate a trajectory for the vehicle as the vehicle moves in the geographic area.