Multi-Sensor Map Alignment Using Shared Pose Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing map generation techniques using single types of localization sensors result in independent maps that are not readily interchangeable, leading to inaccuracies and limitations in localization due to the inherent constraints and characteristics of each sensor type, such as LiDAR and image sensors.
Innovation Solution
A method is employed to align and combine LiDAR and visual sensor data using a shared pose graph optimization, correlating constraints from both data types to generate a unified map with a common coordinate frame, allowing localization with either sensor type.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If maps are generated independently using each sensor type (LiDAR, visual), then each map can be optimized for its specific sensor constraints, but the maps cannot be accurately aligned or registered with each other due to different coordinate frames and warping constraints
Solution Approach 1:
The patent merges multiple independent sensor maps (LiDAR, visual) into a unified map structure by establishing a common coordinate frame. This is achieved through shared pose estimation that correlates sensor data across different types, allowing accurate alignment and registration while preserving the advantages of each sensor type.
Solution Approach 2:
The unified map structure serves multiple sensor types simultaneously, creating a universal representation that can be used for localization with any sensor type. The common coordinate frame and shared pose data enable the map to function as a multi-purpose reference for both LiDAR and visual sensors.
2Adaptability or versatility
If a unified map with common coordinate frame is created, then map interchangeability and multi-sensor utility are improved, but the complexity of aligning and registering different sensor data increases
Solution Approach 1:
The patent introduces shared pose estimation as an intermediary mechanism that mediates between different sensor data types. This shared pose data acts as a common reference that simplifies the alignment process by providing pre-computed transformation relationships between sensor frames, reducing the overall complexity of registering multiple sensor maps.
Solution Approach 2:
The system performs preliminary pose estimation and coordinate frame alignment during the map building process itself, rather than requiring complex post-processing registration. By establishing the common coordinate frame and shared poses upfront during data collection, the system eliminates the need for difficult offline registration operations.
3Measurement precision
If shared pose data is determined using constraints from both sensor types, then localization accuracy is improved, but the computational requirements and processing time increase
Solution Approach 1:
The patent applies partial action by selectively using constraints from sensor data only when they contribute meaningfully to pose estimation. Rather than processing all possible constraints from both LiDAR and visual sensors equally, the system identifies and utilizes the most informative subset of constraints, reducing computational overhead while maintaining accuracy.
Data Source
AI summary
Examples disclosed herein may involve a computing system that is operable to (i) receive first data of one or more geographical environments from a first type of localization sensor, (ii) receive second data of the one or more geographical environments from a second type of localization sensor, (iii) determine constraints from the first data and the second data, (iv) determine shared pose data associated with both of the first data and the second data using the constraints determined from both the first data and the second data by determining one or more sequences of common poses between respective poses generated from each of the first and second data, wherein the shared pose data provides a common coordinate frame for the first data and the second data, and (v) generate a map of the one or more geographical environments using the determined shared pose data.