Vehicle Horizon Object Contextualization for Relevant Roadway Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional object detection systems in vehicles inefficiently process objects that have little to no impact on vehicle operation, leading to unnecessary computational resource usage, as they rely on predefined radii and geometrical shapes, failing to contextualize objects relative to the vehicle's trajectory and road topography.
Innovation Solution
A method that identifies a vehicle horizon based on its trajectory and roadway data, assigning objects to roadway segments within this horizon, allowing for the monitoring and prioritization of relevant objects, thereby reducing unnecessary processing of irrelevant objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional predefined radius methods are used to establish object relevance, then all objects within a fixed distance are processed, but computational resources are wasted on irrelevant objects that have no impact on vehicle operation
Solution Approach 1:
The patent applies local quality by making the object relevance assessment context-dependent rather than uniform. Instead of using a single predefined radius for all objects, the system evaluates each object's relevance based on its specific location, type, and potential impact on vehicle operation. This allows the system to process only those objects within the detection radius that are actually relevant to vehicle safety and operation, eliminating waste of computational resources on irrelevant objects while maintaining reliable detection of important ones.
2Device complexity
If objects are filtered based on distance alone, then processing is simplified, but objects on unrelated roadways or infrastructure are incorrectly included as relevant
Solution Approach 1:
The patent applies segmentation by dividing the detection space into meaningful categories based on object characteristics and location context. Objects are segmented into relevant and irrelevant groups based on multiple criteria including whether they are on the current roadway, their type (vehicle, pedestrian, infrastructure), and their potential impact on vehicle operation. This segmentation approach maintains relatively simple processing while significantly improving relevance classification accuracy by correctly identifying and excluding objects on unrelated roadways or infrastructure.
3Reliability
If all detected objects are monitored, then no relevant object is missed, but processing time and computational load increase unnecessarily
Solution Approach 1:
The patent applies partial action by monitoring only the subset of objects that are relevant to vehicle operation rather than all detected objects. The system performs a preliminary relevance assessment to identify objects that require full monitoring, and excludes irrelevant objects from continuous tracking. This ensures that no relevant object is missed while significantly reducing processing time and computational load by focusing resources only on objects that could impact vehicle safety or operation.
Data Source
AI summary
The present disclosure relates to systems, devices and methods for contextualizing objects relative to a vehicle horizon. In one embodiment, a method is provided including identifying a vehicle horizon based on vehicle trajectory and roadway data relative to the vehicle position and vehicle trajectory. The vehicle horizon is identified to determine at least one roadway segment. Objects relative to the vehicle may be identified and assigned to a roadway segment. Objects may be contextualized by assigning only objects that are associated with the vehicles horizon. Each object assigned to a roadway segment is assigned based on the vehicle horizon. The method may include monitoring each object assigned to a roadway segment associated with the vehicle horizon. By monitoring assigned objects, objects of little to no consequence to the vehicle can be ignored and computing resources of the vehicle may not be wasted on tracking objects of no interest.