Vehicle Lane Type Detection Using Internal Sensor Probabilities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining road types traveled by a motor vehicle are invasive, require significant data, and do not provide precise usage information, lacking flexibility in parameter choice and necessitating extensive retraining.
Innovation Solution
A non-invasive method using internal vehicle sensors to measure parameters like speed, acceleration, and steering angle, calculating probability coefficients for road types based on threshold comparisons, allowing real-time or post-drive adjustments without additional equipment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS systems and navigation data are used to determine vehicle usage, then precise usage information can be obtained, but personal data privacy is compromised
Solution Approach 1:
The invention extracts only the necessary dynamic parameters (acceleration, speed, steering angle) from the vehicle's sensor data while deliberately excluding any geolocation or navigation system data. This extraction approach obtains sufficient usage information (road type, driving conditions) without capturing personal data that would compromise user privacy.
Solution Approach 2:
The invention introduces dynamic vehicle parameters as intermediary indicators that indirectly reflect road type and usage conditions without directly exposing personal data. Instead of using GPS coordinates or navigation data as direct indicators of usage, the system uses acceleration patterns, speed variations, and steering angles as mediators to infer road characteristics.
2Adaptability or versatility
If neural network methods are used to categorize road types, then classification can be achieved, but the system requires significant computational resources and fixed parameters that need extensive retraining
Solution Approach 1:
The invention changes the approach from fixed neural network parameters to dynamic, adaptable thresholds for acceleration, speed, and steering angle. These thresholds can be adjusted based on different road types and driving conditions without requiring complete system retraining, enabling flexible adaptation to new scenarios.
Solution Approach 2:
The invention introduces dynamic thresholds and weighting factors that can adapt to different driving conditions and road types in real-time. Instead of static classification parameters, the system dynamically adjusts its criteria based on the pattern of sensor data, allowing versatile adaptation without extensive retraining.
3Measurement precision
If additional sensors and equipment are installed to improve measurement accuracy, then more precise data can be collected, but device complexity and cost increase
Solution Approach 1:
The invention makes existing vehicle sensors serve multiple functions: acceleration sensors not only monitor vehicle dynamics for safety systems but also provide data for road type classification. Steering angle sensors serve both steering control and road type determination. This multi-functionality approach improves measurement precision without adding dedicated sensors for each function.
Solution Approach 2:
The invention leverages the vehicle's existing sensor infrastructure to serve the additional function of road type determination. The vehicle's own dynamic parameters, already being measured for basic operation, are repurposed to classify road types, allowing the system to improve measurement precision without requiring additional specialized equipment.
Data Source
Figure 1~2
Figure 3~4
AI summary
The invention relates to a method for determining the type of lane used by a motor vehicle, comprising the steps of: - acquiring by a computer, during at least one working interval, the values of at least two parameters which are taken from measurements made by internal sensors equipping the motor vehicle and which are related to the dynamics of the vehicle, - determining by the computer, for each parameter and for the same sampling period included in each of said working intervals, a probability coefficient that the vehicle is on one type of lane rather than another, depending on the values taken by the parameter during said sampling period, and - deducing by the computer, depending on the probability coefficients determined, the type of lane used by the motor vehicle during said sampling period.