Multi-target multi-camera tracking plays a crucial role in intelligent traffic surveillance systems due to its broad application domain. This study investigates the integration and performance of the YOLOv9 object detector and ByteTrack multi-object t...
Multi-target multi-camera tracking plays a crucial role in intelligent traffic surveillance systems due to its broad application domain. This study investigates the integration and performance of the YOLOv9 object detector and ByteTrack multi-object tracker within a multi-camera vehicle tracking framework. Designed to enhance vehicle tracking across various camera views in urban settings, YOLOv9 detects and locates vehicles within frames, while ByteTrack links these detections across frames to establish continuous object trajectories. Experiments on the CityFlow dataset demonstrate noticable improvements in tracking performance compared to baseline methods.