3D LiDAR SLAM With Multi-Resolution Feature Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing LIDAR systems struggle with efficient real-time simultaneous localization and mapping in dynamic urban environments, particularly in autonomous vehicles, due to high computational demands and limitations in constructing three-dimensional maps and locating the LIDAR system within the environment.
Innovation Solution
The method involves segmenting and clustering LIDAR image frames to remove redundant data before feature detection, using elevation and optical properties to identify relevant pixels, and employing multiple resolution SLAM to improve computational efficiency while maintaining mapping and localization accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LIDAR systems process complete high-resolution point cloud data in real-time for SLAM, then mapping and localization accuracy is improved, but computational processing time increases significantly
Solution Approach 1:
The patent segments the complete point cloud data into multiple resolution levels, processing only relevant portions at high resolution while using lower resolution data for less critical areas. This segmentation allows the system to maintain accurate mapping and localization where needed while reducing overall computational processing time.
Solution Approach 2:
The patent applies local quality by assigning different processing resolutions to different spatial regions based on their importance. Critical regions for SLAM (such as areas with features important for localization) receive high-resolution processing, while less critical regions use lower resolution processing, thereby optimizing the balance between accuracy and processing time.
2Loss of information
If LIDAR systems capture complete 3-D point cloud data across broad field of view, then environmental mapping completeness is improved, but data processing complexity increases
Solution Approach 1:
The patent segments the broad field of view into multiple regions of interest and processes each region with appropriate resolution. This maintains environmental mapping completeness by ensuring all areas are captured, while reducing processing complexity by applying different processing strategies to different segments rather than uniformly processing the entire dataset.
Solution Approach 2:
The patent applies partial action by processing only the necessary portions of the point cloud data at full resolution. Instead of uniformly processing all data points with maximum detail, the system selectively applies high-resolution processing only where needed for accurate SLAM, reducing overall processing complexity while maintaining mapping completeness.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach reduces data points by up to 40% and enhances computational efficiency, allowing for real-time, accurate mapping and localization in dynamic environments, suitable for autonomous vehicles.
Implementation Method 1
LIDAR systems employ pulses of light to measure distance to an object based on the time of flight (TOF) of each pulse of light
Implementation Method 2
The light pulses are focused through a lens or lens assembly
Implementation Method 3
Some LIDAR systems employ a single laser emitter/detector combination combined with a rotating mirror to effectively scan across a plane
Data Source
AI summary
Methods and systems for improved simultaneous localization and mapping based on 3-D LIDAR image data. In one aspect, LIDAR image frames are segmented and clustered before feature detection to improve computational efficiency while maintaining both mapping and localization accuracy. Segmentation involves removing redundant data before feature extraction. Clustering involves grouping pixels associated with similar objects together before feature extraction. In another aspect, features are extracted from LIDAR image frames based on a measured optical property associated with each measured point. The pools of feature points comprise a low resolution feature map associated with each image frame. Low resolution feature maps are aggregated over time to generate high resolution feature maps. In another aspect, the location of a LIDAR measurement system in a three dimensional environment is slowly updated based on the high resolution feature maps and quickly updated based on the low resolution feature maps.