Polyline Tracking via Virtual Box Merging for ADAS
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle tracking systems face challenges in reliably determining and processing polylines, especially due to temporal disappearance of objects caused by changes in orientation and movement, leading to unstable detection of driving path boundaries.
Innovation Solution
The method involves assigning virtual boxes along polyline segments, merging them with previous cycles' boxes, and using count values or weight factors to compensate for disappearing elements, generating a convex hull visible from a predetermined viewpoint to stabilize polyline tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If radar sensors are used for sensing the environment and identifying objects, then objects can be detected based on signal reflection, but the quality of identified objects may depend on the reflection properties of the object causing objects to appear or disappear depending on orientation
Solution Approach 1:
The patent merges virtual boxes from multiple detection cycles to create a more reliable representation of objects. By combining detection results across time, the system compensates for objects that temporarily disappear due to orientation changes, improving both reliability and measurement precision of object identification.
Solution Approach 2:
The system performs preliminary actions by creating and storing virtual boxes for detected objects before they may disappear. These pre-established virtual boxes serve as a reference for subsequent detection cycles, allowing the system to maintain object tracking even when temporary detection failures occur due to orientation changes.
2Stability of the object's composition
If virtual boxes are assigned along polyline segments and merged with previous cycles, then polyline tracking stability is improved, but processing complexity increases due to merging operations
Solution Approach 1:
The patent segments the polyline into multiple segments and assigns virtual boxes to each segment independently. This segmentation allows the merging operation to work on smaller, manageable units rather than processing the entire polyline at once, reducing processing complexity while maintaining tracking stability.
Solution Approach 2:
The system creates virtual box copies from previous detection cycles and merges them with current detections. This copying approach allows efficient comparison and merging operations without requiring complex real-time processing, as the virtual boxes can be pre-generated and stored for later merging operations.
Data Source
Figure 1~2b
Figure 3a~3d
Figure 3e~3h
AI summary
The invention relates to a tracking of polylines. A current polyline is processed by assigning virtual boxes to segments of a current polyline and merging these virtual boxes with virtual boxes relating to polylines of a previous processing task. In this way, a stable and reliable tracking of polylines can be achieved. The polylines may be used, for example, for advanced driver assistance systems or a fully automated driving of a vehicle.