Autonomous Map Localization Using Static-Dynamic Object Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in generating accurate and efficient map data for autonomous systems using sensor data from RADAR and LIDAR sensors, particularly in distinguishing between static and dynamic objects.
Innovation Solution
The system processes sensor data to determine whether objects are static or dynamic, and adjusts the map data accordingly, using a point cloud engine to generate RADAR point clouds, a compression engine to reduce data size, and a localization engine to determine pose parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensor data from RADAR and LIDAR sensors is processed to generate map data, then the map data can be used for autonomous navigation, but the data processing complexity and computational requirements increase significantly
Solution Approach 1:
The system segments map data into static objects (unchanging over time) and dynamic objects (changing over time). This segmentation allows different processing strategies to be applied to each type, reducing overall computational complexity while maintaining navigation accuracy.
Solution Approach 2:
The patent extracts and removes dynamic objects from the map data, keeping only static objects for long-term navigation references. This extraction reduces the amount of data that needs to be processed and stored, thereby reducing computational requirements while preserving essential navigation information.
2Loss of information
If all sensor data is included in map data, then complete environmental representation is achieved, but data storage requirements and processing time increase
Solution Approach 1:
Dynamic objects are extracted and removed from the map data, keeping only static objects. This reduces the volume of data to be stored and processed while retaining the essential static environmental structure needed for navigation.
Solution Approach 2:
The system discards dynamic object data from the persistent map storage, as these objects change frequently and are better handled through real-time sensor processing. The static environmental information is recovered and retained for long-term navigation purposes.
3Measurement precision
If dynamic objects are included in map data, then real-time environmental accuracy is improved, but map data stability and consistency deteriorate
Solution Approach 1:
The system segments objects into static and dynamic categories. Static objects form the stable baseline map structure, while dynamic objects are handled separately through real-time detection, maintaining both map stability and environmental accuracy.
Solution Approach 2:
Dynamic objects are extracted from the persistent map data structure. This separation maintains map data stability by excluding frequently changing elements, while dynamic object detection continues to provide real-time environmental accuracy through separate processing channels.
Data Source
AI summary
One or more embodiments of the present disclosure relate to generation of map data. In these or other embodiments, the generation of the map data may include determining whether objects indicated by the sensor data are static objects or dynamic objects. Additionally or alternatively, sensor data may be removed or included in the map data based on determinations as to whether it corresponds to static objects or dynamic objects.