Vehicle Position Estimation Using Road Marker Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional vehicle position determination systems using edge images from cameras have limited accuracy, as they rely on comparing two-dimensional edge data without specific marker detection, which can lead to inaccuracies in estimating the position and attitude of autonomous vehicles.
Innovation Solution
A system that includes a camera, marker detection device, storage device, and controller to capture images, detect specific markers like white lines, and compare them to stored map data, increasing accuracy by assigning higher likelihood to matching pixels, thereby enhancing position determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If edge images from cameras are used for position determination, then the system can operate with conventional camera equipment, but the accuracy of position estimation is limited
Solution Approach 1:
The patent introduces road markers (white lines, stop lines, crosswalks) as intermediary objects between the camera and the position determination process. These markers serve as mediators that provide distinct, easily detectable features in the environment, enabling more accurate position estimation without requiring complex camera systems. The markers act as a bridge between the simple camera input and the precise position output.
2Measurement precision
If specific markers like white line markers or stop line markers are used instead of general edge detection, then position determination accuracy increases, but the complexity of marker detection and processing increases
Solution Approach 1:
The patent applies local quality by focusing detection efforts on specific, locally-defined marker types (white lines, stop lines, crosswalks) rather than attempting to detect all possible edge features. Each marker type has distinct local characteristics (color, shape, position relative to road) that simplify the detection process while maintaining high accuracy. The system tailors its detection approach to the specific local features of road markers rather than using generic edge detection.
3Reliability
If pixels of extracted markers from an image are matched with markers from a virtual map, then the likelihood of correct position determination increases, but the computational processing time increases
Solution Approach 1:
The patent extracts and isolates specific marker features from the entire image for matching purposes, rather than processing all pixels or edge information. By extracting only the relevant marker characteristics (position, type, orientation) and matching these against the virtual map, the system maintains high reliability while reducing the computational burden. This extraction approach focuses processing resources on the most informative features.
Data Source
AI summary
A system for determining a position of a vehicle includes a camera, a marker, a storage device, an external parameter system, and a controller. The camera is configured to capture an image of an area adjacent the vehicle, the image including an edge. The marker detection device is configured to detect a marker in the area adjacent the vehicle. The storage device is configured to store map data, the stored map data including edge data. The positioning system is configured to determine the location of the vehicle relative to the stored map data. The controller is configured to combine the marker detected by the marker detection device and the edge in the image captured by the camera, and compare the combined marker detected by the marker detection device and the edge in the image captured by the camera to the stored map data.