Subpixel-Precise Tracking of Rigid Objects in Real-time

TL;DR

This paper introduces a real-time, subpixel-precise object tracking method that accurately tracks rigid objects using edge-based features, robust to occlusion and illumination changes, and capable of self-diagnosing failures.

Contribution

The paper presents a novel edge-based tracking approach that achieves subpixel accuracy at 80fps, with robustness to occlusion and illumination variations, and includes self-diagnosis of tracking failures.

Findings

Outperforms state-of-the-art real-time trackers in accuracy

Operates at around 80 frames per second

Effectively handles occlusion and illumination changes

Abstract

We present a novel object tracking scheme that can track rigid objects in real time. The approach uses subpixel-precise image edges to track objects with high accuracy. It can determine the object position, scale, and rotation with subpixel-precision at around 80fps. The tracker returns a reliable score for each frame and is capable of self diagnosing a tracking failure. Furthermore, the choice of the similarity measure makes the approach inherently robust against occlusion, clutter, and nonlinear illumination changes. We evaluate the method on sequences from rigid objects from the OTB-2015 and VOT2016 dataset and discuss its performance. The evaluation shows that the tracker is more accurate than state-of-the-art real-time trackers while being equally robust.