AVPS Marker Localization Using Multi-Camera Distortion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting and recognizing coded markers in automated valet parking systems (AVPS) face challenges such as low accuracy due to distance and distortion issues, which affect the precise estimation of the vehicle's position.
Innovation Solution
A method and apparatus that utilize multiple cameras to recognize corner points of coded markers, calculate their coordinates, and estimate the distance by accounting for distortions, thereby improving the accuracy of vehicle localization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single camera is used to detect coded markers, then the device complexity is low, but the measurement precision of vehicle position is insufficient
Solution Approach 1:
The patent combines multiple cameras to form a unified detection system. The coordinate conversion unit transforms coordinates from different camera systems into a unified coordinate system, enabling the system to leverage multiple cameras for improved measurement precision while managing the complexity through systematic coordinate transformation and integration.
2Area of stationary object
If the distance between camera and marker is large, then the detection range is expanded, but the measurement precision of marker position decreases
Solution Approach 1:
The patent transitions from two-dimensional image coordinates to three-dimensional spatial coordinates through coordinate conversion. By calculating real-world positions from multiple camera views and transforming coordinates between different reference frames, the system achieves accurate 3D localization that compensates for distance-related precision loss in 2D images.
3Measurement precision
If conventional marker detection methods are used, then the processing speed is fast, but the measurement precision is low due to distortion effects
Solution Approach 1:
The patent applies distortion correction by transforming marker coordinates from distorted image space to corrected real-world space. The system changes the parameter representation of marker positions from raw pixel coordinates to corrected spatial coordinates, compensating for lens distortion and improving measurement precision while maintaining processing efficiency through established coordinate transformation algorithms.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure provides a method and an apparatus for vehicle localization using AVPS marker. According to at least one embodiment, the present disclosure provides a method for performing localization by a vehicle including a plurality of cameras, including recognizing corner points of a coded marker from an image acquired from each of the cameras, calculating a position, a width, and an angle of the coded marker by using the coordinates of the corner points, filtering the coded marker, calculating a final weight of the coded marker, generating a position and an angle of a local coded marker, rotationally transforming the position of the local coded marker by using information on markers in a parking facility and the angle of the local coded marker, and measuring the position of the vehicle by using the position of the local coded marker and the final weight.