แนวคิดหลัก
Event-based cameras in Intelligent Transportation Systems benefit from fusion with RGB cameras for improved detection performance.
บทคัดย่อ
Event-based cameras offer high temporal resolution and dynamic range, suitable for ITS.
Fusion with RGB cameras enhances color and texture information.
Targetless calibration approach developed to handle multiple moving objects.
Three fusion methods introduced: Early Fusion, Simple Late Fusion, Spatiotemporal Late Fusion.
TUMTraf Event Dataset published with synchronized event-based and RGB images.
สถิติ
During our extensive experiments, we verified the effectiveness of our calibration method with multiple moving objects.
We increased the detection performance by up to +9% mAP in the day and up to +13% mAP during the challenging night.
คำพูด
"Event-based cameras are slowly being established in roadside ITS."
"Our targetless calibration approach can handle multiple moving objects."