Sparse Lane Map Compression for Autonomous Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating efficiently due to the sheer volume of data required for traditional mapping technologies, which limits their ability to process and store map data effectively, and they need optimized solutions for constructing and transmitting sparse maps for navigation.
Innovation Solution
The system employs cameras to analyze images and process data from GPS, sensors, and other sources to construct and navigate using a crowdsourced sparse map, allowing for efficient data transfer and storage by recognizing road features and generating polynomial representations of road segments, enabling accurate vehicle localization and navigation with reduced data requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional mapping technology is used to provide detailed map data for autonomous navigation, then navigation accuracy is improved, but data storage and transmission requirements increase significantly
Solution Approach 1:
The patent extracts only the essential navigation-critical features from complete map data, creating sparse maps that contain only lane markings, road geometry, and key landmarks needed for autonomous navigation. This extraction approach maintains navigation accuracy while dramatically reducing data volume by eliminating redundant information such as detailed building structures, vegetation, and non-essential road features.
Solution Approach 2:
The patent segments map data into discrete lane segments and road features rather than storing continuous detailed maps. Each lane segment is represented by key geometric parameters and sparse feature points, allowing the system to reconstruct navigation-relevant information on-demand while minimizing stored data volume.
2Reliability
If detailed map data is stored and transmitted for autonomous vehicle navigation, then navigation reliability is improved, but bandwidth consumption and processing time increase
Solution Approach 1:
The system extracts only navigation-critical information from complete map datasets, storing and transmitting only sparse representations of roads and lanes. This extraction maintains reliability for navigation tasks while reducing data processing time by minimizing the volume of data that must be loaded, parsed, and analyzed by autonomous vehicle systems.
Solution Approach 2:
The patent applies partial action by providing only the subset of map data necessary for safe autonomous navigation rather than complete environmental detail. The sparse maps include sufficient information for lane keeping, path planning, and obstacle avoidance while omitting excessive detail that would increase processing overhead.
Data Source
AI summary
A system for identifying features of a roadway traversed by a host vehicle may include at least one processor programmed to: receive a plurality of images representative of an environment of the host vehicle; recognize in the plurality of images a presence of a road feature associated with the roadway; determine a location of a first point associated with the road feature relative to a curve representative of a path of travel of the host vehicle; determine a location of a second point associated with the road feature relative to the curve; and cause transmission to a server remotely located from the host vehicle of a representation of a series of points associated with locations along the road feature. The second point may be spaced apart from the first point, and the series of points may include at least the first point and the second point.