High Definition Map Update Using Camera Trajectory and Surface Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Advanced Driver Assistance Systems (ADAS) require precise high definition maps for safe and accurate vehicle control, but existing methods for updating these maps are costly and time-consuming, and struggle with accurately estimating lane marking positions based on surface information.
Innovation Solution
A method and apparatus for updating high definition maps using a camera-mounted vehicle to acquire two-dimensional images, check moving trajectories, generate local landmark maps by estimating three-dimensional positions of lane markings, and update the maps in real-time, leveraging surface information and machine learning algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to update high definition maps, then map accuracy can be maintained, but the cost and time required for updating increase significantly
Solution Approach 1:
The system enables vehicles to automatically update high definition maps using their own cameras and processors during normal operation. The vehicle-mounted camera captures images, the processor generates local landmark maps, and updates are transmitted to the server without requiring dedicated update operations or personnel, thus reducing both time and cost while maintaining accuracy
Solution Approach 2:
The map updating process occurs continuously during vehicle operation rather than through periodic dedicated updates. Vehicles constantly capture road images and generate landmark maps during normal driving, enabling continuous accumulation of map data that reduces overall update time while maintaining high accuracy through ongoing refinement
2Measurement precision
If traditional methods are used to update high definition maps, then comprehensive map coverage can be achieved, but the cost and resources required increase significantly
Solution Approach 1:
Vehicles equipped with cameras serve dual purposes: performing their primary transportation function while simultaneously acting as mobile mapping devices. The same camera used for general vehicle operation is utilized to capture road images for map updating, eliminating the need for specialized expensive equipment and reducing overall system cost while maintaining map accuracy
Solution Approach 2:
The system uses standard camera technology to create accurate visual copies of road surfaces and lane markings. By capturing images with conventional cameras and processing them through image recognition algorithms, the system generates precise landmark maps without requiring expensive specialized surveying equipment, thus reducing costs while maintaining measurement precision
3Measurement precision
If lane marking positions are estimated without using surface information, then processing speed may be faster, but estimation accuracy decreases
Solution Approach 1:
The system pre-processes and stores surface information of roads in the high definition map before needing to estimate lane marking positions. By having surface characteristics (curvature, slope, texture)预先 available in the map data, the system can quickly reference this information during lane marking estimation without performing complex real-time analysis, thus improving accuracy while managing processing complexity
Solution Approach 2:
Surface information acts as an intermediary between the captured images and the lane marking position estimation. The pre-stored surface characteristics of the road serve as a reference framework that guides the interpretation of image data, enabling more accurate lane marking detection by providing contextual information about road geometry without requiring complex direct image analysis
Data Source
AI summary
In accordance with an aspect of the present disclosure, there is provided a method of updating a high definition map. The method comprises, acquiring two-dimensional images at a plurality of different locations by using a camera mounted on a vehicle; checking a moving trajectory of the camera for acquiring the two-dimensional images; generating a local landmark map by estimating, based on surface information of a road in the high definition map, a three-dimensional position of a landmark for a lane marking around the moving trajectory of the camera; and updating the high definition map based on the local landmark map.