Vehicle Collision Detection Using Height-Based Obstacle Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle collision detection systems face inaccuracies and delays due to sensor data, leading to potential collisions, especially when differentiating between obstacles based on height, which compromises driving safety.
Innovation Solution
A method and system that divide obstacle information into height-based maps, model vehicles accordingly, and fuse these models to create collision detection models, allowing for accurate real-time collision risk analysis by determining distance relationships between vehicles and obstacles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor data is used for collision detection, then collision detection can be performed, but the detection accuracy and timeliness are insufficient due to data inaccuracy and delays
Solution Approach 1:
The patent segments the collision detection system into multiple independent detection modules: sensor data acquisition module, image data acquisition module, depth information calculation module, and collision risk determination module. Each module processes specific types of data independently, then integrates results to improve both accuracy and timeliness of collision detection.
Solution Approach 2:
The patent introduces depth information as an intermediary parameter that bridges sensor data and collision detection results. By calculating depth information from image data and combining it with sensor data, the system achieves more accurate and timely collision detection than using sensor data alone.
2Adaptability or versatility
If traditional sensor-based collision detection is used, then general collision detection is possible, but it cannot differentiate between obstacles of different heights leading to inaccurate detection for specific vehicle types
Solution Approach 1:
The patent applies local quality by processing different regions of the image data differently based on obstacle height characteristics. The system divides the detection field into multiple depth regions and applies different detection thresholds and algorithms for each region, enabling accurate detection for both low obstacles (like curbs) and high obstacles (like other vehicles).
Solution Approach 2:
The patent adds the height dimension to traditional 2D collision detection by calculating depth information from image data and combining it with sensor data. This transforms the detection space from 2D to 3D, enabling the system to differentiate obstacles by height and provide vehicle-type-specific collision detection.
3Device complexity
If a unified collision detection model is used for all vehicles, then the system is simple to implement, but it cannot account for different vehicle heights and dimensions
Solution Approach 1:
The patent implements a dynamic detection model that automatically adjusts detection parameters based on the current vehicle type and obstacle characteristics. The system selects different detection thresholds, depth ranges, and fusion weights dynamically according to the vehicle's height and sensor configuration, providing vehicle-type adaptability without requiring multiple fixed models.
Solution Approach 2:
The patent changes detection parameters (such as depth thresholds, fusion weights, and detection regions) based on vehicle type and obstacle height. By adjusting these parameters dynamically, the system adapts to different vehicle configurations while maintaining a unified detection framework, balancing complexity and versatility.
Data Source
AI summary
The present application provides a vehicle collision detection method and system and a vehicle. The vehicle collision detection method includes: dividing an acquired collection of obstacle information into a first obstacle map and a second obstacle map based on a preset obstacle height; modeling a current vehicle based on the first obstacle map and the second obstacle map, and fusing modeling results with the first obstacle map and the second obstacle map to obtain a first collision detection model and a second collision detection model; and further, obtaining a distance relationship between a preset position of the current vehicle and obstacles corresponding to the collection of obstacle information through the first collision detection model and/or the second collision detection model, and determining whether the current vehicle is in collision based on the distance relationship.