Feature-Based Navigation Map Generation Using Camera Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation maps lack essential road features such as width, humps, intersection angles, and slope, requiring expensive hardware like LIDAR scanners and continuous server connections, which is computationally intensive and prone to disruptions.
Innovation Solution
A method and system for generating and updating vehicle navigation maps by extracting and correlating these features using sensors, reducing the need for continuous server connections and expensive hardware, by processing environmental data to create feature-based maps that can be stored and shared for subsequent use.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If expensive hardware like LIDAR scanners is used to extract road features, then measurement precision of road features is improved, but device complexity and cost increase
Solution Approach 1:
The patent uses cameras to capture images of the road environment, creating a visual copy of the physical road features. These images are then processed to extract road features such as width, humps, intersections, and slopes, replacing the need for expensive LIDAR scanners while maintaining measurement capability through image analysis
Solution Approach 2:
The patent replaces mechanical scanning systems (LIDAR) with optical systems (cameras). Instead of using mechanical or electromagnetic scanning to measure road features directly, the system uses image capture and computer vision algorithms to detect and extract road features from visual data
2Measurement precision
If dynamic extraction of road features is performed using vehicle sensors, then navigation map accuracy is improved, but computational load increases
Solution Approach 1:
The patent extracts and stores road feature data in advance during normal vehicle operation and communication periods. This preliminary extraction and storage of features like road width, humps, and intersections reduces the need for intensive real-time computation during autonomous navigation, as the data is already prepared and available
Solution Approach 2:
The system performs feature extraction selectively based on communication availability and vehicle operation conditions. Instead of continuously extracting all possible features at maximum intensity, the system adjusts the extent of extraction based on available resources, performing partial extraction when needed and storing results for later use
3Loss of information
If continuous connection with central server is maintained for map updates, then navigation information currency is improved, but reliability decreases due to connection interruptions
Solution Approach 1:
The patent divides the navigation system into independent segments: vehicles that extract and transmit road feature data, central servers that store and manage the data, and receiving vehicles that access the data. This segmentation allows the system to function with intermittent connections, as each segment can operate independently when disconnected
Solution Approach 2:
Vehicles in the patent perform self-service by extracting road feature data during their own operation and transmitting it to the network. This distributed approach means the system does not rely on a single centralized data collection point, and vehicles can contribute to and benefit from the collective knowledge base even when disconnected from the central server
Data Source
AI summary
This disclosure relates generally to vehicle navigation maps, and more particularly to method and system for generating and updating vehicle navigation maps with features of navigation paths. In one embodiment, a method may include receiving a position of a vehicle and an environmental field of view (FOV) of the vehicle along a navigation path of the vehicle on a navigation map, extracting features of the navigation path from the environmental FOV, correlating the features with the navigation path on the navigation map based on the position, generating a features based navigation map based on the correlation, and transmitting the features based navigation map to a server of a navigation map service provider for storage and for subsequent use. The features based navigation map, when required to assist a navigation of another vehicle, may be accessed to assess the features of the navigation path and to provide the assessment to the other vehicle.