Automatic Parking Path Segmentation for Low-Cost Accurate Guidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic parking assistance systems are costly due to expensive sensors and controllers, and they require improved guiding accuracy.
Innovation Solution
An in-vehicle device with a communication interface receives a parking navigation path from an external device, which is divided into linear and arc road sections, allowing the device to control the vehicle's behavior based on line-shape features, determining if it can travel along each section and adjusting the path if necessary, with deviation thresholds to ensure accurate parking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If expensive sensors and controllers are used for automatic parking assistance, then guiding accuracy is improved, but system cost increases
Solution Approach 1:
The parking navigation path is segmented into multiple road sections with different line-shape features (linear, arc, etc.). Each section is analyzed independently to determine vehicle travel feasibility, allowing for more precise and cost-effective control compared to treating the entire path as a single unit.
Solution Approach 2:
The system pre-computes the parking navigation path and divides it into road sections before the actual parking maneuver. This preliminary path planning and segmentation allows the vehicle to follow a predetermined accurate path without requiring expensive real-time sensing and computation during execution.
2Measurement precision
If complex path computation is performed in real-time during parking, then guiding accuracy is improved, but computing complexity and time consumption increase
Solution Approach 1:
The parking navigation path computation and road section segmentation are performed in advance before the parking maneuver begins. This shifts the computational burden from real-time execution to pre-processing, significantly reducing the computing complexity and time requirements during the actual parking operation while maintaining high guiding accuracy.
Solution Approach 2:
By segmenting the path into predefined road sections with simple geometric characteristics (linear, arc), the complex continuous path following problem is transformed into a series of simpler discrete section evaluation and execution tasks, reducing real-time computing complexity.
3Reliability
If the vehicle continuously adjusts movements based on real-time sensor tracking, then collision-free parking is achieved, but system cost and device complexity increase
Solution Approach 1:
The system determines in advance whether the vehicle can travel along each road section based on pre-computed line-shape features. This preliminary feasibility assessment allows the vehicle to follow a pre-determined safe path without requiring continuous real-time sensor tracking and adjustment, reducing system complexity while maintaining collision-free operation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The disclosure provides an in-vehicle device, a system and a method for automatic parking assistance. The in-vehicle device comprises a communication interface through which the in-vehicle device receives a parking navigation path from an external device, the parking navigation path comprising linear road sections and/or arc road sections; and a parking controller coupled with the communication interface. The parking controller is configured to identify line-shape features of each road section; determine whether the vehicle can travel along a road section based on line-shape features of the road section; in the case that the determination is affirmative, control a driving behavior of the vehicle such that the vehicle travels along the road section; in the case that the determination is negative, determine a new travelling path and control a driving behavior of the vehicle such that the vehicle travels along the new travelling path.