Neural Network Parking Space Detection in Navigation Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation systems for autonomous and semi-autonomous vehicles lack accurate parking space detection, leading to potential parking violations and operational inefficiencies, especially under varying lighting conditions.
Innovation Solution
A navigation system utilizing a multilayer neural network to analyze sensor data, identify real-world endpoints and bounding boxes, apply line rules, and generate an overhead depiction of parking spaces, merging maneuvering instructions for safe and reliable parking operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional parking detection methods are used, then the system is simpler to implement, but the accuracy of parking space detection deteriorates under varying lighting conditions
Solution Approach 1:
The patent replaces traditional mechanical/optical parking detection methods with a neural network-based system. The neural network model processes sensor data to identify parking spaces, replacing conventional sensors and algorithms that fail under varying lighting conditions. This substitution enables accurate detection across different lighting environments while maintaining system feasibility through software-based processing.
Solution Approach 2:
The patent changes the operational parameters of the detection system by using a neural network that can adapt to varying lighting conditions. The system processes sensor data through trained neural network models that have learned to recognize parking space patterns regardless of illumination levels, effectively changing how the system responds to environmental parameter variations.
2Reliability
If accurate parking space detection is implemented, then parking operation reliability improves, but the complexity of data processing increases
Solution Approach 1:
The patent applies preliminary action by pre-training neural network models with extensive sensor data before deployment. The models are trained in advance to recognize various parking space patterns and lighting conditions, so that during actual parking operations, the system can quickly and reliably identify parking spaces without requiring complex real-time processing. This pre-processing of knowledge enhances reliability while keeping operational complexity manageable.
3Measurement precision
If multilayer neural network analysis is used, then parking space identification accuracy improves, but computational time increases
Solution Approach 1:
The neural network models are trained in advance with extensive sensor data covering various lighting conditions and parking space configurations. This pre-training allows the models to make accurate predictions during actual parking operations without requiring complex real-time computations, thereby reducing computational time while maintaining high identification accuracy.
Data Source
AI summary
A navigation system includes: a user interface configured to: receive a sensor data packet for a scan area; a control circuit, coupled to the user interface, configured to: analyze the sensor data packet submitted to a multilayer neural network already trained including storing the sensor data packet; parse the sensor data packet with the multilayer neural network to generate real world endpoints and a bounding box including identifying the real world coordinates; apply line rules, including identifying a parking space by the real world endpoints including storing the real world endpoints in the storage circuit; compile an overhead depiction including boundary lines of the parking space identified by the real world endpoints and the bounding box in the scan area; merge vehicle maneuvering instructions into the overhead depiction for accessing the parking space; and present the overhead depiction for displaying on a user display.