Achieving Adaptive Visual Multi-Object Tracking with Unscented Kalman Filter

As an essential part of intelligent monitoring, behavior recognition, automatic driving, and others, the challenge of multi-object tracking is still to ensure tracking accuracy and robustness, especially in complex occlusion environments. Aiming at the issues of the occlusion, background noise, and...

Full beskrivning

Bibliografiska uppgifter

Huvudupphov:	Guowei Zhang, Jiyao Yin, Peng Deng, Yanlong Sun, Lin Zhou, Kuiyuan Zhang
Materialtyp:	Artigo
Språk:	Inglês
Utgiven:	MDPI AG 2022-11-01
Serie:	Sensors
Ämnen:	multi-object tracking YOLOv5 object detection improved DeepSORT unscented Kalman filter adaptive algorithm
Länkar:	https://www.mdpi.com/1424-8220/22/23/9106
Taggar:	Lägg till en tagg Inga taggar, Lägg till första taggen!