Redundant Road Lane Detection Clustering for Map Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation systems face challenges with inaccurate and low-quality sensor data, particularly in road lane detection, due to incorrect sensor detection, redundancies, and noise, which can lead to hazardous navigation issues in autonomous or semi-autonomous vehicles.
Innovation Solution
A system and method for identifying and removing redundant road lane detections by clustering sensor data based on criteria such as matched distance, parallel detections, and heading differences, using a processor to establish groups and update map data, ensuring accurate navigation guidance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If sensor data is collected from multiple sources to improve coverage, then the quantity of detection data increases, but redundant and incorrect detections increase
Solution Approach 1:
The patent combines multiple sensor detections by establishing clusters of road lane detections from different sensors and time points. The system merges overlapping detections into unified detection groups, eliminating duplicates while preserving valid detections. This is achieved through clustering algorithms that group detections based on spatial proximity and temporal coherence.
Solution Approach 2:
The system changes detection parameters by evaluating detections against multiple criteria including parallel-detection criterion (comparing lane marking orientations), heading-difference criterion (comparing vehicle heading with detection orientation), and temporal coherence. Detections that fail these parameter evaluations are identified as redundant and removed.
2Loss of information
If all sensor detections are retained to ensure complete coverage, then no detection information is lost, but navigation accuracy decreases due to redundant data
Solution Approach 1:
The patent implements a selective discarding mechanism where redundant detections are identified and removed based on evaluation criteria. The system discards detections that are parallel to existing detections, have inconsistent headings, or fail temporal coherence checks. This selective discarding maintains detection completeness for valid observations while removing harmful redundancies.
Solution Approach 2:
The system employs feedback mechanisms by continuously evaluating detections against established criteria and updating the detection set iteratively. Each detection is evaluated in the context of previously accepted detections, and the results feed back into the detection set for the next evaluation cycle, progressively improving accuracy.
3Measurement precision
If redundant detections are removed to improve data quality, then navigation accuracy improves, but the complexity of data processing increases
Solution Approach 1:
The patent segments the detection processing into distinct modular stages: initial detection collection, clustering formation, evaluation against multiple criteria (parallel-detection, heading-difference, temporal coherence), and redundant detection removal. Each stage handles a specific aspect of processing, making the overall complex system manageable and efficient.
Solution Approach 2:
The system performs preliminary actions by pre-establishing evaluation criteria and clustering structures before final detection determination. Detections are pre-clustered and pre-evaluated against multiple criteria, so that the final redundant removal step operates on already-processed data, reducing the computational burden of the final decision.
Data Source
AI summary
A system for identifying redundant road lane detections in map data and subsequently updating the map data to remove the redundant road lane detections is provided. The system may be configured to determine, based on sensor data, a plurality of road lane detections associated with a road link represented by the map data. The system is further configured to determine a cluster for the road link based on a clustering criterion. The system is further configured to establish a plurality of road lane detection groups based on connectivity of the road lane detections in the cluster. The plurality of road lane detection groups is evaluated to identify one or more redundant road lane detections based on one or more of a parallel-detection criterion or a heading difference criterion. The identified redundant road lane detections are used to update the map data by a computer-implemented update process.