Parking Space Gradient Estimation Using Key Point Rotation Matrices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for autonomous parking struggle to accurately estimate the gradient of parking spaces, especially when the road surface is sloping, leading to deteriorated autonomous parking control performance.
Innovation Solution
A method and apparatus that estimate the gradient of a parking space by detecting key points corresponding to the parking space through image recognition, obtaining angle information formed by these key points, and calculating a rotation matrix to determine the gradient using a reference rotation matrix.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If camera calibration is performed under the assumption that the road surface is on the same plane, then the gradient estimation is simplified, but the estimation accuracy deteriorates when the parking space exists on a sloping road surface
Solution Approach 1:
The patent changes the parameter of road surface assumption from flat to sloping by introducing gradient estimation. It detects key points (entrance line, end line, parking lines) and calculates rotation matrices to estimate the gradient angle, thereby adapting the calibration process to sloping conditions and improving location estimation accuracy without excessive complexity increase
Solution Approach 2:
The patent performs preliminary gradient estimation before autonomous parking control. By detecting key points in advance and estimating the gradient using rotation matrices, the system prepares accurate parking space location information beforehand, which improves subsequent parking control performance
2Productivity
If a single gradient value is used for all parking spaces, then the processing is simplified, but the autonomous parking control performance deteriorates when parking spaces have different gradients
Solution Approach 1:
The patent segments the gradient estimation process by detecting key points for each parking space (entrance line, end line, parking lines) and estimating gradients individually. This allows different gradient values to be assigned to different parking spaces, improving control reliability while maintaining reasonable processing efficiency through automated key point detection
Solution Approach 2:
The patent applies local quality by estimating gradient specifically for each parking space based on its own key points rather than using a global gradient value. This localized approach ensures that each parking space's unique slope characteristics are captured, improving autonomous parking control accuracy for diverse parking environments
Data Source
AI summary
An apparatus for estimating a parking space gradient includes a parking space detection module that recognizes a type of a parking space based on an image captured by an image capturing device and recognizes a location of the parking space by detecting a plurality of key points corresponding to the parking space and a gradient estimation module that obtains angle information formed by the key points in the type of the parking space, estimates a rotation matrix based on estimation conditions previously set according to the angle information and the type of the parking space, and estimates a gradient of the parking space using the estimated rotation matrix and a predetermined reference rotation matrix.