Vision-Lidar Tunnel Mapping for Poor-Texture UAV SLAM
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional tunnel modeling methods using UAVs face challenges in accurately modeling tunnels with poor texture and dark, narrow areas, leading to inaccuracies in 3D modeling due to limited data collection and high picture repeatability.
Innovation Solution
A method and system integrating a depth camera and lidar for SLAM mapping, utilizing Bayesian fusion to combine point cloud and laser information, and feature matching to improve map model accuracy, with the assistance of positioning and auxiliary UAVs for enhanced data collection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional UAV data collection methods are used in tunnels, then the system is simple and easy to operate, but the modeling accuracy deteriorates in poor texture and dark areas
Solution Approach 1:
The patent combines multiple sensing systems (depth camera, lidar, and positioning systems) into an integrated UAV data collection platform. The depth camera captures texture information while the lidar provides depth data, and their fusion compensates for individual sensor limitations in poor texture environments, thereby improving modeling accuracy without requiring separate complex systems
Solution Approach 2:
The patent introduces auxiliary positioning UAVs as intermediary devices that provide reference signals and lighting support. These intermediary elements facilitate the main UAV's data collection by providing external reference points for positioning and additional illumination for depth camera operation in dark tunnel areas
2Reliability
If only depth camera is used for data collection, then the device is simple, but the data quality deteriorates in dark and narrow areas
Solution Approach 1:
The patent merges depth camera and lidar sensors into a complementary sensing system. The depth camera provides texture and color information while the lidar provides accurate depth measurements, especially in dark areas where the depth camera struggles. Their fused data creates more reliable and comprehensive 3D models
Solution Approach 2:
The integrated sensor system serves multiple functions: the depth camera captures visual texture information for areas with sufficient lighting, while the lidar provides depth data in dark and narrow areas. This multi-functional approach ensures reliable data collection across varying tunnel conditions without requiring separate specialized systems
3Measurement precision
If feature matching is used to correct map models, then the positioning accuracy is improved, but the computational time increases
Solution Approach 1:
The patent performs preliminary feature extraction and matching during the data collection phase, establishing correspondence between depth camera images and lidar point clouds before final map model construction. This preliminary feature matching creates reference relationships that accelerate subsequent positioning and model correction operations
Solution Approach 2:
The patent implements an iterative feedback mechanism where the constructed map model is continuously corrected through feature matching between consecutive frames. The positioning accuracy improvements from feature matching feed back into refined map models, which in turn provide better references for subsequent feature matching operations, creating a progressive refinement cycle
Data Source
AI summary
The method includes: obtaining point cloud information collected by a depth camera, laser information collected by a lidar, and motion information of an unmanned aerial vehicle (UAV); generating a raster map based on the laser information, and obtaining pose information of the UAV based on the motion information; obtaining a map model through fusing the point cloud information, the raster map, and the pose information by a Bayesian fusion method; and correcting a latest map model by feature matching based on a previous map model.