Reverse Driving Assistance With Position Correction From Stored Paths
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle reversing systems face challenges in accurately estimating vehicle location due to location estimation errors that accumulate during forward driving, which are difficult to correct in narrow roads or underground parking lots, and existing methods like V-SLAM require excessive computation.
Innovation Solution
A driving assistance method and device that store a driving path during forward driving, correct vehicle position using sensor, GPS, and feature information with weighted adjustments, and compensate for GPS delays, utilizing multiple camera views for position correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If V-SLAM is used to correct position during backward driving, then position estimation accuracy is improved, but computation load increases excessively
Solution Approach 1:
The position correction process is segmented into multiple stages: pre-storing map information during forward driving, selecting candidate correction points from stored data, and performing matching only for selected candidates. This divides the computation into manageable segments rather than processing all data continuously, reducing real-time computation load while maintaining accuracy
Solution Approach 2:
Map information and feature points are stored in advance during forward driving before backward driving begins. This preliminary action prepares correction data beforehand, so that during backward driving, the system only needs to perform matching operations on pre-stored data rather than acquiring and processing new data in real-time, significantly reducing computation load
2Adaptability or versatility
If location estimation is based on internal vehicle sensors, then the system can operate independently, but location estimation errors accumulate during driving
Solution Approach 1:
The system uses stored map information as a reference feedback mechanism. During backward driving, the current sensor-based position estimate is compared against the pre-stored map features, and correction values are computed based on the difference. This feedback loop continuously corrects accumulated errors without requiring external systems
Solution Approach 2:
The system creates a copy of the environment map during forward driving and stores it for later use. This copied map serves as a reference for position correction during backward driving, allowing the system to compare actual sensor data against the stored copy to identify and correct position drift
3Measurement precision
If GPS information is used to correct position, then location accuracy is improved, but it cannot be used in narrow roads or underground parking lots
Solution Approach 1:
The system uses feature-based map matching that can operate in multiple environments including narrow roads, underground parking lots, and open areas. By storing and matching visual features rather than relying on GPS coordinates alone, the system achieves universal applicability across diverse environments where different positioning methods may fail
Data Source
AI summary
A driving assistance, a driving assistance device therefor, and a computer-readable non-volatile storage medium therefor are provided. The method includes storing a driving path of a vehicle based on forward driving of the vehicle, in response to the vehicle driving backward, correcting a position of the vehicle based on at least one of sensor information measured by an internal sensor of the vehicle, global positioning system (GPS) information of the vehicle, or feature information obtained through image information of the vehicle, and assisting backward driving of the vehicle based on the stored driving path and the corrected position of the vehicle.