LiDAR Contour Analysis for Autonomous Vehicle Parking Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face inaccuracies in recognizing the size and direction of parked vehicles due to limitations in ultrasonic sensors and cameras, leading to inaccurate parking space estimation, which increases the number of parking control steps and degrades marketability.
Innovation Solution
A parking control device and method utilizing a LiDAR sensor to obtain contour information of parked vehicles, identify parking types, estimate vehicle sizes, and explore target parking spaces, thereby improving accuracy and reducing the number of parking control steps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ultrasonic sensors and cameras are used for parking space recognition, then the system can detect parked vehicles, but the accuracy of parking space recognition is insufficient due to ground clearance error and sensor limitations
Solution Approach 1:
The patent introduces an intermediary computational model that uses visible area ratios and contour information as intermediate representations to bridge the gap between sensor data and accurate parking space recognition. This intermediary processing layer compensates for the limitations of direct sensor measurements.
Solution Approach 2:
The patent changes the parameters used for detection from direct distance measurements to visible area ratios and contour-based parameters. By transforming the detection parameters, the system achieves more accurate parking space recognition that is independent of ground clearance variations.
2Measurement precision
If LiDAR is used to supplement parking space recognition, then accuracy improves, but invisible areas cannot be accurately detected
Solution Approach 1:
The patent performs preliminary actions by estimating the size and direction of parked vehicles before attempting to detect invisible areas. This preliminary estimation provides a framework that guides subsequent detection efforts and improves the accuracy of invisible area identification.
Solution Approach 2:
The patent implements feedback mechanisms where the estimated vehicle parameters are continuously refined based on LiDAR data and contour information. This feedback loop allows the system to improve its detection of invisible areas through iterative optimization.
3Productivity
If vehicle size and direction are estimated only using sensor input values, then the process is simple, but the estimation accuracy is insufficient leading to increased parking control steps
Solution Approach 1:
The patent adds another dimension to the estimation process by incorporating contour information and visible area ratios alongside traditional sensor data. This multi-dimensional approach significantly improves estimation accuracy without substantially increasing system complexity.
Solution Approach 2:
The patent creates a composite estimation model that combines multiple data sources (sensor inputs, contour information, visible area ratios) into a unified estimation framework. This composite approach leverages the strengths of each data source to achieve high accuracy in vehicle size and direction estimation.
4Device complexity
If inaccurate estimation of target vehicle size and direction is used, then the system can operate with simple sensors, but the parking space exploration becomes inaccurate increasing control steps
Solution Approach 1:
The patent performs preliminary estimation of vehicle parameters using simple sensor data before conducting detailed parking space exploration. This preliminary action provides accurate initial parameters that guide the exploration process, reducing the number of iterative control steps required.
Solution Approach 2:
The patent transforms simple sensor inputs into accurate vehicle parameter estimates through parameter transformation using contour information and visible area ratios. This parameter change enables accurate parking space exploration without requiring complex sensor systems.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution enhances the accuracy of parking space recognition and vehicle size estimation, reducing the number of parking control steps and improving the marketability and reliability of autonomous vehicles by accurately determining the size and direction of target vehicles and parking spaces.
Implementation Method 1
using a Light Detection and Ranging (LiDAR) sensor to obtain contour information of a parked vehicle
Data Source
AI summary
A method of controlling parking of a vehicle provided with a parking control device including a processor includes: under the control of the processor, obtaining first contour information from a first parked vehicle which is parked in a parking space using sensor information obtained via a sensor module provided in the vehicle; identifying a plurality of parking types for the first parked vehicle based on the first contour information and a current traveling direction of the vehicle; estimating a size of the first parked vehicle in response to a parking type selected among the plurality of parking types; and exploring a target parking space in the parking space based on the size of the first parked vehicle.