Autonomous Driving Map Generation via Spatial Submap Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing high-precision mapping technologies for autonomous driving are hindered by a heavy computation burden due to dependence on initial values of point clouds, leading to poor scenario adaptability and slow mapping optimization, especially in weak GPS environments.
Innovation Solution
The method involves acquiring global initial poses of multiple point clouds using a multi-circle collection mode, dividing them into spatial submap graphs based on distribution relationships, optimizing these poses within each submap graph, and stitching the submap graphs together to generate a base map, thereby reducing computational load and improving stitching efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If point cloud data is processed with initial value estimation and optimization iteration to generate the bottom layer of the high-precision map, then the map accuracy is improved, but the computation burden increases and mapping optimization cannot be accomplished quickly
Solution Approach 1:
The patent divides the point cloud data into multiple spatial submap graphs based on spatial distribution relationships. Each submap graph is processed independently through optimization, which reduces the overall computational complexity compared to processing all point clouds as a single large dataset. This segmentation enables faster mapping optimization while maintaining accuracy.
2Measurement precision
If heavy computation is performed in the optimization process to achieve accurate mapping, then the mapping precision is improved, but the time consumption increases
Solution Approach 1:
By segmenting the optimization process into multiple spatial submap graphs, the patent reduces the time required for each optimization iteration while maintaining overall mapping precision. The divided submaps can be processed in parallel or sequentially with reduced computational overhead compared to a monolithic optimization approach.
3Device complexity
If the system depends heavily on initial values of point clouds for base graph generation, then the initial processing is simplified, but the scenario adaptability deteriorates in weak GPS environments
Solution Approach 1:
The patent segments the point cloud processing into spatial submap graphs that can be independently optimized with less dependence on global initial values. This segmentation enables the system to adapt to various scenarios including weak GPS environments, as each submap can be processed with local optimization rather than relying heavily on global initial value estimates.
Data Source
AI summary
The present application discloses a high-precision mapping method and device, which relates to the field of autonomous driving. A specific implementation includes: acquiring global initial poses of multiple point clouds, where the point clouds are point clouds of a location for which a map is to be built and are collected by a lidar using a multi-circle collection mode; dividing the multiple point clouds into multiple spatial submap graphs according to a spatial distribution relationship of the multiple point clouds; optimizing, for each spatial submap graph, global initial poses of point clouds belonging to the spatial submap graph to acquire global poses of the point clouds in each spatial submap graph; and stitching the multiple spatial submap graphs together according to global poses of the point clouds in the multiple spatial submap graphs to acquire a base graph of the map to be built.