High-Precision Electronic Map Construction via Grid Map Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for constructing high-precision electronic maps, particularly in urban environments with weak or missing geolocation signals, suffer from large cumulative errors and low efficiency due to inaccurate initial pose of lidar devices, which cannot meet the high precision requirements.
Innovation Solution
A method involving the acquisition of multiple point cloud sequences for a preset region, transformation into grid maps, and optimization through grid map matching to construct high-precision point cloud maps, even in scenarios with weak or missing geolocation system signals, by combining accurately measured and locally smoothed poses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If high-precision electronic maps are constructed using lidar point cloud data, then map accuracy is improved, but the system becomes highly sensitive to initial pose accuracy and geolocation signal quality
Solution Approach 1:
The patent introduces grid maps as an intermediary representation between raw point cloud data and the final electronic map. By transforming point cloud data into grid maps with reflection value information, the system creates a robust intermediate form that can be matched and optimized even when initial pose data is inaccurate or geolocation signals are weak, thus mediating between the high precision requirement and the unreliable environment
Solution Approach 2:
The patent implements a feedback mechanism by optimizing point cloud sequences based on grid map matching results. The system repeatedly refines the point cloud data by comparing grid maps from different acquisitions and adjusting the point cloud sequences to improve consistency, creating a closed-loop feedback process that enhances reliability without sacrificing precision
2Manufacturing precision
If multiple point cloud sequences are acquired and optimized through grid map matching, then map precision is improved, but processing time and computational complexity increase
Solution Approach 1:
The patent segments the point cloud data into multiple sequences from different acquisition turns and processes them through grid map representation. By dividing the large-scale point cloud data into manageable segments that can be independently processed and then matched, the system achieves high precision through multiple sequences while making the processing time more manageable through efficient grid-based operations
Solution Approach 2:
The patent changes the representation parameters of the point cloud data by transforming it into grid maps with specific reflection value information. This parameter transformation allows for more efficient comparison and matching operations, enabling the system to process multiple point cloud sequences with improved precision while reducing the computational burden compared to direct point cloud matching
Data Source
AI summary
Embodiments of the present disclosure provide a method and a device for generating an electronic map, an electronic device, a computer readable storage medium, and an acquisition entity. The method includes: obtaining a first point cloud sequence and a second point cloud sequence for a preset region; generating a first grid map for the first point cloud sequence and a second grid map for the second point cloud sequence, wherein a grid in each of the first grid map and the second grid map at least comprises reflection value information of a point cloud; and optimizing the first point cloud sequence based on the first grid map and the second grid map.