3D Roadway Map Generation by Filtering Moving Object Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in generating accurate 3D maps of roadways due to distortions caused by moving objects, which can interfere with navigation systems.
Innovation Solution
A method and device that process sensor data from lasers to create a 2D grid of the roadway, identify the lowest relative elevation for each cell, and generate a 3D map by removing data points exceeding a threshold distance above this elevation, thereby excluding data from moving objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If sensor data from all detected objects is used to generate the 3D map, then the map contains complete information about the environment, but moving objects distort the map and reduce navigation accuracy
Solution Approach 1:
The patent extracts and removes data points corresponding to moving objects from the sensor data before generating the 3D map. By identifying and separating moving object data from stationary environment data, the system maintains completeness of environmental information while eliminating distortions that would reduce navigation accuracy.
Solution Approach 2:
The patent segments the sensor data into different categories based on object motion characteristics. By dividing the data points into moving objects and stationary environment elements, the system can selectively process each segment differently, keeping complete environmental information while removing harmful moving object distortions.
2Manufacturing precision
If all data points from sensor data are processed to create detailed 3D maps, then the map resolution and detail are improved, but processing time and computational complexity increase
Solution Approach 1:
The patent extracts only the necessary data points for map generation by removing moving object data early in the processing pipeline. This extraction reduces the total number of data points that require detailed processing, thereby maintaining high map resolution while reducing computational time and complexity.
Solution Approach 2:
The patent performs preliminary filtering of moving objects from the sensor data before the main 3D map generation process. By removing unnecessary data points in advance, the system reduces the computational burden of subsequent processing steps, achieving both high resolution and efficient processing time.
3Productivity
If data points from moving objects are included in the 3D map, then the sensor data is fully utilized, but the reliability of the map for navigation purposes decreases
Solution Approach 1:
The patent extracts and removes moving object data points from the sensor data before generating the 3D map. This ensures that only reliable stationary environment data is used for navigation, eliminating the negative impact of moving objects on map reliability while still utilizing available sensor data efficiently.
Solution Approach 2:
The patent converts the potentially harmful presence of moving objects in sensor data into a benefit by using their detection as a criterion for data filtering. By identifying moving objects and excluding them from map generation, the system improves navigation reliability while maintaining efficient use of sensor data.
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 allows for the generation of accurate 3D maps that exclude moving objects, enhancing navigation accuracy and reliability for autonomous vehicles.
Implementation Method 1
receiving sensor data collected by a laser
Data Source
AI summary
Aspects of the present disclosure relate generally to safe and effective use of autonomous vehicles. More specifically, laser data including locations, intensities, and elevation data may be collected along a roadway in order to generate a 3D map of the roadway. In order to remove extraneous objects such as moving vehicles from the 3D map, a 2D grid of the roadway may be generated. The grid may include a plurality of cells, each representing an area of the roadway. The collected data may be sorted into the grid based on location and then evaluated to identify the lowest elevation of the cell. All data points above some threshold distance above this lowest elevation may be removed. The resulting data points may be used to generate a 3D map of the roadway which excludes the extraneous objects.