Forward Collision Avoidance Object Filtering by Driving Environment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current forward collision avoidance systems (FCA) face unnecessary computational load and malfunctions due to misrecognition of objects, particularly in varying driving environments, such as highways, without considering the specific types of objects present.
Innovation Solution
A method and apparatus that collect driving environment information, select candidate objects based on road type, and verify object recognition to reduce computational load and prevent malfunctions by excluding unsuitable objects, using a reference table and cross-checking sensor data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the FCA performs an operation to select target objects for each object type without considering the driving environment, then the system can detect all possible objects, but it causes unnecessary computational load
Solution Approach 1:
The patent segments the object detection process by dividing objects into different types (pedestrian, cyclist, vehicle, animal) and applying environment-specific filtering to each type. The candidate object selection unit selectively determines which object types to track based on the driving environment, thereby reducing computational load while maintaining detection completeness for relevant objects.
Solution Approach 2:
The system dynamically adjusts the set of candidate object types based on the driving environment. The candidate object selection unit changes which object types are tracked in real-time according to environmental conditions (e.g., highway vs. urban area), making the detection process adaptive and efficient rather than static and computationally expensive.
2Adaptability or versatility
If the FCA selects target objects for each object type without considering the driving environment, then the system can identify all object types, but it causes misrecognition of objects
Solution Approach 1:
The patent applies local quality by making the object detection strategy specific to each driving environment. Different environments (highway, urban, rural) have different expected object types, and the system adjusts its detection focus locally to each environment, improving recognition accuracy by not forcing detection of irrelevant object types in certain contexts.
Solution Approach 2:
The system performs preliminary action by pre-determining which object types are relevant based on the driving environment before actual detection occurs. The candidate object selection unit提前 identifies which object types to look for based on environmental context, preventing misrecognition by avoiding detection attempts for object types that are unlikely to be present in the current environment.
3Adaptability or versatility
If the FCA checks all object types in all environments, then the system can handle diverse scenarios, but it increases the complexity of the system operation
Solution Approach 1:
The patent implements universality by creating a unified FCA system that can handle multiple driving environments and object types through a single adaptive framework. The candidate object selection unit serves as a universal mechanism that adjusts candidate object types based on environment, allowing one system to perform multiple environment-specific functions without requiring separate detection systems for each scenario.
Data Source
AI summary
A method for forward collision avoidance assist includes: collecting driving environment information of a vehicle; detecting at least one front object positioned in front of the vehicle and acquiring front object information of the front object; generating a driving trajectory of the vehicle based on motion information of the vehicle; selecting a candidate object from the front object based on the driving environment information and the front object information; determining a target object based on information of the candidate object and the driving trajectory of the vehicle; and controlling a brake system included in the vehicle based on a risk of collision between the vehicle and the target object.