Vehicle Detection and Tracking From Surveillance Cameras in Urban Scenes

TL;DR

This paper presents a vehicle detection and tracking system for urban scenes that enhances multi-object tracking by integrating re-identification features, improving accuracy during occlusions and fast movements, while maintaining real-time processing speed.

Contribution

It introduces an extension of IOU-based tracking with re-identification features to better handle occlusions and fast motions in urban vehicle tracking.

Findings

Outperforms baseline on UA-DETRAC benchmark

Maintains real-time processing speed

Improves tracking accuracy during occlusions

Abstract

Detecting and tracking vehicles in urban scenes is a crucial step in many traffic-related applications as it helps to improve road user safety among other benefits. Various challenges remain unresolved in multi-object tracking (MOT) including target information description, long-term occlusions and fast motion. We propose a multi-vehicle detection and tracking system following the tracking-by-detection paradigm that tackles the previously mentioned challenges. Our MOT method extends an Intersection-over-Union (IOU)-based tracker with vehicle re-identification features. This allows us to utilize appearance information to better match objects after long occlusion phases and/or when object location is significantly shifted due to fast motion. We outperform our baseline MOT method on the UA-DETRAC benchmark while maintaining a total processing speed suitable for online use cases.