Vehicle Map Fusion for Dynamic Object Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomously operated vehicles face challenges in accurately distinguishing dynamic objects from static features in the environment, leading to inefficient map generation and increased processing resources.
Innovation Solution
A method and system that combines information from vehicle sensors, including cameras and radar, to create a fused map by identifying dynamic objects through velocity analysis, using occupancy grids to filter out moving objects and enhance map accuracy for autonomous navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If sensor systems capture all objects in the environment to create a comprehensive map, then the map completeness is improved, but dynamic objects are mistakenly represented as static features which worsens the map accuracy
Solution Approach 1:
The patent segments the mapping process into multiple stages: initial map creation using all sensor data, followed by a separate dynamic object identification and removal phase. This segmentation allows the system to maintain comprehensive object detection while subsequently improving accuracy by eliminating false static representations of dynamic objects.
Solution Approach 2:
The patent extracts dynamic objects from the comprehensive map through velocity analysis and temporal comparison. By identifying and removing dynamic object representations, the system maintains map completeness for static features while improving accuracy by eliminating erroneous static representations of moving objects.
2Loss of information
If the map includes all detected objects for comprehensive navigation information, then the navigation information completeness is improved, but processing resources are increased which worsens computational efficiency
Solution Approach 1:
The patent extracts only the necessary navigation-relevant static features from the complete sensor data, removing dynamic objects that do not contribute to path planning. This extraction maintains navigation information completeness for static environmental features while improving computational efficiency by reducing the data volume requiring processing.
Solution Approach 2:
The patent discards dynamic object data from the navigation map after it has been used for collision avoidance detection, recovering computational resources. The system maintains awareness of dynamic objects through separate tracking while removing them from the static navigation map, thus preserving navigation completeness while improving processing efficiency.
3Manufacturing precision
If the system processes all sensor data to create detailed maps, then the map detail precision is improved, but the time required for map generation is increased which worsens real-time performance
Solution Approach 1:
The patent performs preliminary velocity analysis and dynamic object identification during data acquisition phases, before final map generation. By pre-identifying dynamic objects through velocity thresholds and temporal comparison, the system reduces the computational burden during map creation, thus maintaining detailed precision while improving real-time performance.
Solution Approach 2:
The patent extracts dynamic object candidates early in the processing pipeline using velocity-based filtering, removing them from subsequent detailed mapping operations. This early extraction maintains map detail precision for static features while reducing overall processing time by excluding dynamic objects from intensive mapping computations.
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
Enhances map accuracy by removing dynamic objects, reducing processing resources, and enabling efficient, collision-free navigation for autonomous vehicles.
Implementation Method 1
the object velocities comprise a Doppler velocity
Data Source
AI summary
A method and system for a vehicle control system generates maps utilized for charting a path of a vehicle through an environment. The method performed by the system obtains information indicative of vehicle movement from at least one vehicle system and images including objects within an environment from a camera mounted on the vehicle. The system uses the gathered information and images to create a depth map of the environment. The system also generates an image point cloud map from images taken with a vehicle camera and a radar point cloud map with velocity information from a radar sensor mounted on the vehicle. The depth map and the point cloud maps are fused together and any dynamic objects filtered out from the final map used for operation of the vehicle.