Paper Title :Weather-Informed Vision Enhancement for Autonomous Vehicles in Adverse Conditions
Author :Emin Bayramov, Zoltan Istenes
Article Citation :Emin Bayramov ,Zoltan Istenes ,
(2024 ) " Weather-Informed Vision Enhancement for Autonomous Vehicles in Adverse Conditions " ,
International Journal of Advance Computational Engineering and Networking (IJACEN) ,
pp. 1-5,
Volume-12,Issue-8
Abstract : Providing Advanced Driver Assistance Systems (ADAS) features requires high-quality image data collected by
vehicles. However, adverse weather conditions and nighttime significantly degrade image quality, negatively impacting
object detection accuracy and model performance for ADAS function- alities. This paper addresses this critical issue by
referencing relevant works that have encountered similar challenges. We propose a novel solution that utilizes the vehicle’s
GPS location and data collection timestamp to query weather forecast via a weather API. By obtaining precise weather
details at the timeandlocationofdatacollection,weenhanceimagequalitythrough a pre-processing step tailored to the specific
weather conditions. UsingtheDAWN(DetectioninAdverseWeatherNature)dataset, our approach demonstrates substantial
improvements in image clarity and object detection accuracy across various weather scenarios, significantly enhancing the
robustness and reliabilityof object detection models for ADAS systems.
Keywords - ADAS Features, Image Enhancement, Adverse Weather Conditions, Object Detection, Weather Api
Type : Research paper
Published : Volume-12,Issue-8
DOIONLINE NO - IJACEN-IRAJ-DOIONLINE-21168
View Here
Copyright: © Institute of Research and Journals
|
 |
| |
 |
PDF |
| |
Viewed - 26 |
| |
Published on 2024-11-19 |
|