High definition (HD) maps are one of the key technologies supporting autonomous-driving vehicles (ADVs). Especially in urban scenarios, the field of view of sensors is often limited, and HD map provides critical information about upcoming road environmental data. Maps used for ADVs are high resolution with centimeter-level accuracy and their correctness is fundamental when analyzing the safety of upcoming maneuvers. This paper proposes an approach for online map validation (OMV) based on spatial and temporal correlation of smart-sensors. Smart sensors are capable of analyzing the validity of regions of the map independently from one another. Results from the sensors are then fused over multiple regions and time samples for providing a unified view to software components deciding on upcoming maneuvers which areas of the maps are consistent with sensor data and which are not.

Correlation-based approach to online map validation

Cenedese A.
2020

Abstract

High definition (HD) maps are one of the key technologies supporting autonomous-driving vehicles (ADVs). Especially in urban scenarios, the field of view of sensors is often limited, and HD map provides critical information about upcoming road environmental data. Maps used for ADVs are high resolution with centimeter-level accuracy and their correctness is fundamental when analyzing the safety of upcoming maneuvers. This paper proposes an approach for online map validation (OMV) based on spatial and temporal correlation of smart-sensors. Smart sensors are capable of analyzing the validity of regions of the map independently from one another. Results from the sensors are then fused over multiple regions and time samples for providing a unified view to software components deciding on upcoming maneuvers which areas of the maps are consistent with sensor data and which are not.
2020
IEEE Intelligent Vehicles Symposium, Proceedings
978-1-7281-6673-5
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11577/3378101
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 3
  • ???jsp.display-item.citation.isi??? ND
social impact