Lidar coarse edge correlation for rapid ground navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inertial navigation systems suffer from error accumulation over time, leading to location and attitude drift, while LIDAR systems generate large datasets that are time-consuming to process for determining current location, especially during initialization.
Innovation Solution
A navigation system that employs a LIDAR sensor to generate both fine and coarse grid representations of environmental data, correlating current data with historical data using edge detection and data fusion with inertial and GNSS data to rapidly determine location, reducing drift and processing time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LIDAR systems generate large quantities of detailed environmental data, then measurement precision and representation detail are improved, but processing time and computational complexity increase
Solution Approach 1:
The patent segments the environmental data into multiple grid representations with different levels of detail (coarse grids and fine grids). The coarse grids contain down-sampled data for rapid processing, while fine grids preserve detailed information. This segmentation allows the system to process coarse grids quickly for initial localization and only process fine grids in detail when necessary, thus reducing overall processing time while maintaining measurement precision.
Solution Approach 2:
The patent applies partial processing by only processing a subset of the LIDAR data at full resolution. Specifically, it processes coarse grids at reduced resolution and only processes fine grids in specific regions where high precision is needed. This partial action approach processes less data than the full dataset would require, significantly reducing processing time while maintaining necessary measurement precision for navigation.
2Measurement precision
If LIDAR data is correlated against historical data to determine current location, then navigation accuracy is improved, but initialization time increases significantly
Solution Approach 1:
The patent segments the correlation process into multiple stages using different grid resolutions. First, coarse grids are correlated with historical coarse data to quickly identify a candidate location region. Then, fine grids are correlated with historical fine data within that restricted region to determine the precise location. This segmented approach reduces initialization time by limiting the search space in the second stage while maintaining high location determination accuracy.
Solution Approach 2:
The patent performs preliminary correlation using coarse grids before conducting the final precise correlation with fine grids. This preliminary action using lower-resolution data quickly narrows down the possible location range, preparing the system for the subsequent detailed correlation step. This preliminary processing reduces the computational burden of the final accurate location determination, significantly reducing overall initialization time.
3Productivity
If inertial navigation systems integrate acceleration and rotation rates over time, then location and velocity information are provided, but error accumulation causes location drift
Solution Approach 1:
The patent implements feedback by periodically correlating current LIDAR grid data with historical LIDAR data to determine the actual location, then using this determined location to correct the inertial navigation system's estimated location. This feedback loop continuously monitors and corrects the drift that accumulates in inertial navigation over time, maintaining location accuracy while allowing the INS to provide continuous navigation solutions.
Solution Approach 2:
The patent creates a composite navigation system that combines inertial navigation with LIDAR-based correlation navigation. The inertial navigation provides continuous high-frequency position estimates, while the LIDAR correlation provides periodic absolute position corrections. This composite approach leverages the strengths of both systems to maintain both continuous navigation provision and long-term location accuracy.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A method includes generating current coarse edge count representation based on current fine grid representation of current section, correlating current edge quantity values of current coarse pixels with historical edge quantity values of historical coarse pixels of historical coarse edge count representation of environment, and identifying first subsection of historical coarse edge count representation with highest correlation to current coarse edge count representation. Each current coarse pixel in current coarse edge count representation represents current fine pixels from current fine grid representation. Fine grid representation of current section of environment is based on data from range and attitude sensor. Each current coarse pixel within current coarse edge count representation includes current edge quantity value that represents quantity of current fine pixels represented by current coarse pixel that include edge. Each historical coarse pixel corresponds to historical fine pixels in historical fine grid representation of environment.