Autonomous Vehicle Interface Trajectory Overlap Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing user interfaces for autonomous vehicles face complexity when displaying multiple predicted trajectories of road agents and ego-vehicles from a driver's perspective view, leading to cluttered and confusing displays, especially when trajectories overlap with objects.
Innovation Solution
A system and method that includes a trajectory-prediction module to generate and modify predicted trajectories for a vehicle user interface, determining if road agent trajectories exist behind objects and overlap with them, and updating the interface to remove or modify these overlaps, thereby simplifying the display by hiding, diluting, or blending trajectories to improve clarity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If multiple road agent trajectories are displayed on the user interface, then the information completeness is improved, but the display complexity increases and visual clutter occurs
Solution Approach 1:
The patent segments the display of trajectories by spatial relationships - separating trajectories that are occluded by objects from those that are visible. This segmentation allows the system to manage multiple trajectories without displaying all of them simultaneously in a cluttered manner, thus maintaining information completeness while reducing display complexity.
Solution Approach 2:
The patent introduces a new dimension to trajectory display by adding occlusion status information. Trajectories are not only displayed based on their spatial position but also their visibility relationship with objects. This dimensional addition allows the system to convey more information about multiple trajectories without increasing visual clutter, as the occlusion dimension provides contextual understanding.
2Loss of information
If all predicted trajectories are displayed from driver's perspective view, then the driver's understanding of road agent positions is improved, but visual clutter and confusion increase due to overlapping trajectories
Solution Approach 1:
The patent extracts and separately handles occluded trajectory information from the main display. By identifying trajectories that are blocked by objects and treating them differently (e.g., hiding them or displaying them with reduced prominence), the system removes the harmful visual clutter effect while preserving the underlying information about road agent positions for the driver's understanding.
Solution Approach 2:
The patent uses visual differentiation through color or transparency changes to distinguish between visible and occluded trajectories. This allows the driver to understand the spatial relationships and occlusion status of different trajectories without the overlapping lines creating confusion. The visual property change helps separate overlapping information in a comprehensible way.
Data Source
AI summary
Systems and methods are provided for generating trajectories for a vehicle user interface showing a driver's perspective view. Methods include generating an ego-vehicle predicted trajectory for an ego-vehicle; and generating at least one road agent predicted trajectory for a road agent that is external to the ego-vehicle. After the predicted trajectories are generated, the method continues by determining that at least one road agent predicted trajectory both (1) exists behind an object, indicating travel behind the object, and (2) overlaps with the object, when displayed on the user interface showing a driver's perspective view. The method includes modifying the at least one road agent predicted trajectory to remove the overlap. The method then proceeds with updating a display of the user interface to include any modified road agent predicted trajectory. Systems include a trajectory-prediction module to execute the methods.