Video-Based Vehicle Positioning and Mapping With Pixel-Wise Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation and mapping systems for autonomous vehicles face challenges in accurately localizing vehicles within road networks, integrating data from diverse sensors, and maintaining up-to-date pre-built maps, while efficiently handling computational and bandwidth limitations.
Innovation Solution
A method and system that utilizes camera-based image processing to generate a local map representation of a vehicle's environment, which is compared to a reference map to determine geographical location and orientation, and updates the reference map with new data, incorporating techniques like visual odometry and semantic segmentation to enhance accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If ad-hoc sensing is used for map generation, then the map can be created without pre-built data, but sensing errors accumulate and localization precision deteriorates
Solution Approach 1:
The system performs preliminary actions by using the vehicle's odometry system to pre-determine its own position and orientation before comparing sensor data with the pre-built map. This preliminary localization step provides an accurate reference frame for subsequent map generation, preventing error accumulation while maintaining adaptability to new environments.
2Measurement precision
If pre-built maps are used for localization, then localization accuracy is improved, but the maps become outdated and require frequent updates
Solution Approach 1:
The system implements feedback by continuously comparing odometry-based position estimates with locations derived from pre-built map features. This feedback loop enables real-time detection of map outdatedness and triggers selective updates only in areas where changes are detected, maintaining reliability while reducing unnecessary update overhead.
Solution Approach 2:
The mapping system transitions from a static pre-built map approach to a dynamic hybrid system that combines pre-built map data with real-time odometry-based updates. This dynamic approach allows the map to adapt to changes in the environment while maintaining the structural benefits of pre-built maps for localization.
3Manufacturing precision
If comprehensive sensor data is collected for map generation, then mapping accuracy is improved, but computational resources and bandwidth consumption increase
Solution Approach 1:
The system extracts and processes only the essential features needed for localization and mapping rather than comprehensively processing all sensor data. By selectively extracting relevant geometric and semantic features from sensor inputs, the system maintains mapping accuracy while significantly reducing computational energy consumption and bandwidth requirements.
4Measurement precision
If multiple sensors are integrated for environment sensing, then sensing accuracy is improved, but system complexity increases
Solution Approach 1:
The system achieves multi-functionality by using the vehicle's odometry system for dual purposes: both for navigation/motion planning and for providing position references in map generation. This universal use of existing sensors improves sensing accuracy without adding the complexity of dedicated multi-sensor integration systems.
Data Source
AI summary
A method and system for determining a geographical location and orientation of a vehicle travelling through a road network is disclosed. The method comprises obtaining, from one or more cameras associated with the vehicle travelling through the road network, a sequence of images reflecting the environment of the road network on which the vehicle is travelling, wherein each of the images has an associated camera location at which the image was recorded. A local map representation representing an area of the road network on which the vehicle is travelling is then generated using at least some of the obtained images and the associated camera locations. The generated local map representation is compared with a section of a reference map, the reference map section covering the area of the road network on which the vehicle is travelling, and the geographical location and orientation of the vehicle within the road network is determined based on the comparison. Methods and systems for generating and/or updating an electronic map using data obtained by a vehicle travelling through a road network represented by the map are also disclosed.