An Occlusion-Aware Multi-Target Multi-Camera Tracking System

Andreas Specker, Daniel Stadler, Lucas Florin, Jurgen Beyerer; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2021, pp. 4173-4182

Abstract


Multi-camera tracking of vehicles on a city-scale level is a crucial task for efficient traffic monitoring. Most of the errors made by such multi-target multi-camera tracking systems arise due to tracking failures or misleading visual information of detection boxes under occlusion. Therefore, we propose an occlusion-aware approach that leverages temporal information from tracks to improve the single-camera tracking performance by an occlusion handling strategy and additional modules to filter false detections. For the multi-camera tracking, we discard obstacle-occluded detection boxes by a background filtering technique and boxes overlapping with other targets using the available track information to improve the quality of extracted visual features. Furthermore, topological and temporal constraints are incorporated to simplify the re-identification task in the multi-camera clustering. We give detailed insights into our method with ablative experiments and show its competitiveness on the CityFlowV2 dataset, where we achieve promising results ranking 4th in Track 3 of the 2021 AI City Challenge.

Related Material


[pdf]
[bibtex]
@InProceedings{Specker_2021_CVPR, author = {Specker, Andreas and Stadler, Daniel and Florin, Lucas and Beyerer, Jurgen}, title = {An Occlusion-Aware Multi-Target Multi-Camera Tracking System}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2021}, pages = {4173-4182} }