Camera-Based Autonomous Parking With External Path Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic valet parking systems require costly and complex installation of environmental sensors outside the vehicle, such as lidar, ultrasonic, and radar sensors, which are difficult to maintain and calibrate, making them impractical for widespread adoption.
Innovation Solution
A method and system utilizing existing camera sensors and processing devices outside the vehicle to capture and process light signals, determine vehicle pose and movement paths, and transmit commands for autonomous driving, leveraging neural networks and edge computing to navigate vehicles safely without additional sensor retrofitting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If environmental sensors (lidar, ultrasonic, radar) are installed outside the vehicle in the entire parking area, then autonomous parking capability is achieved, but system cost and complexity increase significantly
Solution Approach 1:
The patent uses camera sensors to capture optical images of the environment and processes these images to create a digital representation of the parking area. This optical copying approach replaces the need for expensive physical sensor installations (lidar, ultrasonic, radar) throughout the parking facility, achieving autonomous parking capability while dramatically reducing system complexity and cost.
Solution Approach 2:
The patent substitutes mechanical/optical sensing systems (cameras capturing light) for electromagnetic sensing systems (lidar, radar, ultrasonic sensors). By using cameras to capture environmental data and processing these images to determine vehicle positions and navigation paths, the system achieves autonomous parking without requiring complex sensor infrastructure installation in the parking area.
2Reliability
If multiple environmental sensors are procured, mounted, calibrated, and maintained, then safe automated parking is enabled, but operational cost and maintenance burden increase
Solution Approach 1:
The patent replaces multiple types of environmental sensors with camera-based optical sensing. Cameras capture images of the parking environment, and image processing algorithms extract navigation information. This approach enables safe automated parking while eliminating the procurement, mounting, calibration, and maintenance requirements for lidar, ultrasonic, and radar sensors, significantly reducing operational burden.
Solution Approach 2:
The patent makes camera sensors perform multiple functions: capturing environmental geometry for navigation, detecting vehicle positions, identifying parking spaces, and providing obstacle detection. This multi-functionality eliminates the need for separate specialized sensors for each function, reducing both maintenance burden and system complexity while maintaining safety.
3Extent of automation
If production automobiles are equipped with large computational power and numerous sensors, then autonomous driving methods can be deployed, but manufacturing cost increases
Solution Approach 1:
The patent extracts the complex computational processing and environmental sensing functions from the vehicle itself and relocates them to external servers. The vehicle only needs basic camera sensors and communication capabilities to transmit data and receive navigation instructions. This extraction allows standard production automobiles without expensive sensor suites or high-performance onboard computers to achieve autonomous parking capability, significantly reducing manufacturing costs.
Solution Approach 2:
The patent introduces an intermediary communication system (network transmission) between the vehicle and external processing servers. The vehicle captures images with simple cameras and transmits them externally for processing. This intermediary approach allows complex autonomous driving functions to be implemented using minimal onboard equipment, enabling deployment in standard production vehicles without costly modifications.
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
Enables safe and efficient autonomous parking without the need for additional sensor installation, reducing costs and complexity while utilizing existing infrastructure.
Implementation Method 1
capturing the vehicle with the aid of at least one sensor, in particular camera, arranged in the surroundings of the vehicle
Data Source
AI summary
The disclosure relates to a method for remote-controlled autonomous driving of a vehicle (201) having the steps of capturing the vehicle (201) with the aid of at least one sensor, in particular camera, arranged in the surroundings of the vehicle (201), determining a movement path for the vehicle (201) by a processing device (501) located outside the vehicle, transmitting the movement path and/or control commands regarding the movement path to the vehicle (201), and implementing the movement path and/or the control commands in the vehicle (201), in order to move the vehicle (201) in accordance with the movement path.